Contents. Research Articles

Transcription

1 Contents Research Articles Alterations in the Components of the Hedgehog Signaling Pathway Lead to Aberrant Stem Cell Behaviors in the Drosophila Ovary Eva Nong 1 Personality Perception in Toddlers: Trait Awareness in Four-Year-Olds as Related to Threat, Consistency, and Behavior A Prospectus for an Honors Thesis in the Department of Psychology Margaret Arnold 8 Trends Reason and the Passions: From Plato to Neuroeconomics Dmitry Taubinsky 16 Perelman, Poincaré, and the Ricci Flow Scott Duke Kominers 23 Perspectives So Hot Right Now: Environmentalism is All the Rage at Harvard and Beyond Firth McEachern 28 Improving the Access Gap: Biological Expressions Inequalities Gabriella B. Tantillo 31 Robocup 2007 Rishi Gupta 36 Diversity, PRISE, and an Inteview with Professor Robert A. Lue Stephanie Lo 38 Humor PRISE Horoscopes Brandon Geller 40 1

2 Alterations in the Components of the Hedgehog Signaling Pathway Lead to Aberrant Stem Cell Behaviors in the Drosophila Ovary Eva Nong 1 1. Introduction 1.1. Hedgehog signaling. The Hedgehog family of signaling proteins regulates development from flies to humans. [1] The signaling pathway derives its name from the hedgehog protein, a secreted morphogen that activates downstream gene transcription. First discovered in Drosophila, Hedgehog is now known to mediate an array of developmental events like neural tube development and lung branching. Furthermore, aberrant Hedgehog signaling has been implicated in many cancers like basal cell carcinoma, the most common cancer in the Western world. [1] Generally, the pathway is initiated when the hedgehog ligand moves to target cell, where it binds to its receptor Patched (Ptc), a twelve-pass transmembrane protein. Ptc normally inhibits the activity of Smoothened (smo), which resembles G- protein-coupled receptors. When Hedgehog signal is present, Smoothened is released from patched inhibition, ultimately preventing the processing of a zinc-finger transcriptional factor named Cubitus interruptus (Ci) through a complex that includes a kinesin-like protein Coastal 2 (cos2), a serine-threonine protein kinase Fused (Fu) and the Suppressor of Fused (SuFu) [1] See Fig. 1. Full-length Ci (Ci-155) is activates transcription but its binding partners, Cos 2 and Suppressor of Fused (SuFu), limit its access to the nucleus and its specific activity. Hh signaling increases the specific activity of Ci-155 and increase its access to the nucleus, but the majority of Ci-155 still remain in the cytoplasm.the resulting full-length Ci translocates into the nucleus and, with its activity increased with the Hedgehog signal, serves as a transcriptional activator for target genes. In the absence of Hedgehog signal, Ci undergoes ubiquintin-dependent partial proteolysis to a shorter form, Ci-75 that represses transcription of the target genes Stem cell development. Some of the target cells for Hedgehog signals are stem cells. In fact, Hedgehog was one of the first regulators shown to directly affect stem cell proliferation. [4] Understandably, alterations in the pathway may have implications in tumorigenesis. [1, 4] Recently, it has been proposed that certain cancers may be initiated by mutations in tissue-specific stem cells. [8] This is supported by the observation that somatic stem cells and cancer cells share several 1 Eva Nong 09 is a biology concentrator and 2006 PRISE Fellow. Figure 1. The Hedgehog pathway. When the Hedgehog signal is present, patched dissociates from smoothened, which inhibits Cos2. Cos2 regulates the processing of Ci-155 and inhibits its translocation into the nucleus. With Hh signal, Ci remains full-length and translocates into the nucleus, where it serves as a transcription activator for Hh response genes. 2

3 2. MATERIALS AND METHODS 3 essential attributes, such as their abilities to self-renew and differentiate. [5] Mutations in components of the Hh pathway such as Ptc and Smo have been found in basal cell carcinoma. Thus, it is important to study how alterations in the Hh pathway affect stem cell behaviors to promote tumorgenesis. [8] 1.3. The Drosophila ovary. The Drosophila ovary serves as an ideal model to study the Hedgehog pathway in stem cells because it offers some key advantages. First, while there are variations in the hedgehog pathway among different species, the overall signaling scheme is conserved between Drosophila and mammals. [1] Thus, the Drosophila ovary may yield insights into the pathway applicable to other organisms. Secondly, in mammalian tissues, studying the Hedgehog pathway is difficult because it is challenging to correctly identify stem cells because of their scarcity and the absence of molecular markers. The Drosophila ovary addresses the problem because a few identifiable stem cells are located in the middle of a tubular structure, the germarium, from which eggs develop. [2] The ease of identification makes possible mutant studies that examine the effects on stem cells quantitatively and qualitatively. [2] Finally, the short generation time and the sequenced genome of Drosophila allow for the manipulation of Drosophila genotypes. [2] This paper specifically discusses the gene expression and proliferative behaviors of somatic stem cells (SSCs) in response to Hh signal in the Drosophila ovary. Female flies have a pair of ovaries with 12 to 16 ovarioles in each. Each ovariole contains a germarium. Although two to three SSCs are normally positioned in the middle of the germarium, their maintenance is regulated by extracellular signals secreted from neighboring cells anteriorly such as terminal filament and polar cells. [2] The germarium then buds off later stage egg chambers, which receive lower levels of Hh signal because of their distance from the Hh-secreting cells Mutant analysis. SSC clones that are homogyzous for a variety of recessive alleles can be created by FLP-mediated mitotic recombination. [5] This system is induced by heat-shock induction of the FLP recombinase protein. [5] Using this method it has previously been shown that homozygous mutant clones of the positive regulator of Hh signal, Smo, and the adhesion molecule, DE-cadherin (Shg) cause a loss of stem cells. Mutations in the negative regulator Patched, in contrast, exhibit excess stem cell phenotypes. [4] A previous study showed that Patched-null mutants produced an excess of SSC. [4] The number of SSCs quickly doubled in ptc mutants compared to that of wild type ovarioles. [4] The egg chambers also showed signs of excess follicle cells, the progeny of SSCs, which accumulated between the egg chambers. [4] Although some increase in follicle cell number is inevitable as a consequence of increased stem cell numbers, it appears that the excess Hh signaling in SSC progeny also contributes to the excess follicle cell phenotypes. [4] Thus, ptc mutants exhibited phenotypes similar to those with hyper-active Hedgehog stimulation. Ptc mutations also rescued the loss of SSCs normally observed when DE-cadherin is inactivated. The specific goal of this project is to determine the effects of loss-of-function mutations in the other regulators of Hedgehog (cos2, PKA, and Smo) in stem cell maintenance by examining Hh target gene expression and stem cell proliferation. Cos2 is mainly a negative regulator of Hh signal but also has other complex effects, as shown by previous studies that study Hh in the Drosophila wing disc. []10 Without a Hh signal, the microtubule-bound cos2 anchors Ci in the cytoplasm and regulates the processing of Ci. In the presence of Hh, cos2 also serves as the transducer in wing discs. It remains unclear though likely that cos2 acts in a similar manner in the ovary. In a different complex, Su(Fu) regulates the processing and cleavage of Ci. Using the heat-shock inducible flippase FRT system [5], we created SSC lineages homozygous for null mutations in cos2 alleles (cos2-2, W1, P2835B), PKA, and Smo. We then stained and visualized them under the confocal microscope. We also constructed cos2+shg double mutants to see if Cos null mutants can rescue stem cell loss due to DE-cadherin inactivation. Our results show that cos2 mutants exhibited excess follicle cells. Furthermore, cos2 mutants showed a loss of stem cells overtime, confirming that the stem cells themselves are not overproliferating. cos2+shg double mutants showed some rescue of stem cell loss. The failure of Cos mutants to mimic ptc phenotypes possibly indicates that the hedgehog pathway branches off downstream of patched, bypassing Cos. Our results may contribute to our ultimate goal of unveiling insights into the hedgehog pathway and the determinants of stem cell-initiated tumorgenesis. 2. Materials and Methods 2.1. Genetics. As described previously, mutant smo and Su(Fu) clones marked by the absence of GFP stain were derived using standard two-time heat shock of 60 minutes at 37 C. [6] The SSC clones were identified by negative staining through the loss of GFP transgene and their anterior position to Fas-stained follicle cells. The mutant stem cell clones were examined in flies of various genotypes depending on the protein inactivation. See Fig Immunoflorescence. The adult flies were dissected for their ovaries 7 and 14 days post heat shock. The larvae were dissected 10 and 14 days post heat shock. Ovaries were stained as described previously for -galactosidase, Fas III, and Ci, and the secondary antibodies used Alexa fluorescence probes (Molecular Probes). [9] Briefly, ovaries were dissected in 1X PBS and fixed for 20 minutes in paraformaldehyde. The ovaries were permeabilized with 1% Triton X-100 in PBS for 1 hour, blocked in 5% BSA for 1 hour, then incubated with primary secondary for 1 hour in room temperature or at 4C overnight. Ovaries were rinsed with PBST three times for ten minutes each before incubating with secondary antibody for 1 hour in room temperature or at 4C overnight. The antibodies used in this study were anti--galactosidase (1:2000, Cappel), anti-fasciclin III (1:250, Developmental Studies Hybridoma Bank), anti-ci (1:1, R. Holmgren). The secondary were Alexa 488 anti-rabbit, alexa 594 anti-mouse, alexa 647 anti-mouse, and alexa 594 anti-rat antibodies. Microscopy for fluorescent staining was performed on a BioRad MRC-600 confocal microscope system.

4 4 THE HEDGEHOG SIGNALING PATHWAY Table of Fly Genotypes Mutation Genotype FRT recombinase site Smo 2 Ywflp/yw; ubigfpfrt40a/smo 2 FRT40A; ptclacz/+ 40A PKA (DCO B3 ) Ywflp/yw; ubigfpfrt40a/ DCO B3 FRT40A; ptclacz/+ 40A Smo D16 /PKA (DCO B3 ) Ywflp/yw; ubigfpfrt40a/smo 2 DCO B3 FRT40A; ptclacz/+ 40A Control Ywflp/yw; ubigfpfrt40a/cknmfrt40a 40A Cos2 w1 Ywflp/yw; FRT42BubiGFP/FRT42BCos2 w1 ; ptclacz/+ 42B Cos2 P2835b Ywflp/yw; FRT42BubiGFP/FRT42B Cos2 P2835b ; ptclacz/+ 42B Cos2 w1 shg R69 Ywflp/yw; FRT42BubiGFP/FRT42BCos2 w1 ; ptclacz/+ 42B Cos2 w1 shg Ywflp/yw; FRT42BubiGFP/FRT42BCos2 w1 ; ptclacz/+ 42B Ptc Ywflp/yw; FRT42DubiGFP/FRT42Dptc s2 Sha 42D Control Ywflp/yw; FRT42DubiGFP/FRT42DSha 42D Cos2 2 Ywflp/yw; FRT42DubiGFP/FRT42DCos2 2 42D Figure 2. The genotypes of flies that contain the mutant clones. Fly crosses were set up to generate the particular mutation through heat shock-induced FRT recombination. 3. Results 3.1. cos2-inactivated stem cell clones showed elevated ptc expression in later egg chambers. To identify the effects of cos2 inactivation on Hh target gene expression, we examined ptc expression in wild-type and cos2 mutant stem cell lineages by staining with antibodies against -gal. In Drosophila ovary, ptc is the only known Hh response gene. [4] We induced homozygous mutations for different alleles of cos2 (cosw1 and cos2 P2835b) marked cell clones generated by heat shock induced mitotic recombination. Flies were heat shocked as adults and then dissected and examined 7 and 14 days later See Fig. 3. The brightness of the Ptc-lacZ stain was used to evaluate the levels of ptc expression. Figure 3. ptc-lacz expression in Cos2 clones. Adult flies were heat shocked two times in 37C water bath. 7 and 14 days later, they were dissected and stained with anti--gal antibodies to detect ptc-lacz activity. The image in A shows the anti--gal antibody stain for allele Cos2 P2835b and the white arrows indicate the location of the homozygous Cos2- mutant clones. The ovary in B shows the endogenous GFP stain superimposed with anti--gal antibodies in Cos2-w1 clones. The white arrows indicate the Cos2-w1 clones. Our results illustrated the potentially complex role of cos2 in Hh signaling See Tab. 4. While cos2 is a negative regulator by promoting the processing of Ci in the absence of Hh signal, cos2 also serves as a transducer in the presence of Hh signal in the wing discs. For example, because Hh is known to be secreted from cap and terminal filament cells in the anterior end of the germarium, the SSCs in the germarium receive a much higher Hh concentration than follicle cells in the later egg chambers. Thus, Hh response genes and SSC phenotypes are expected to be more active in germarial than follicle cells.

5 3. RESULTS 5 Likewise, the cos2 homozygous mutation may have an inhibitory effect in the germarium because cos2 is no longer available to transducer the Hh signal; whereas in the egg chambers the mutation would have an activating influence because Ci is not processed into Ci-75 even when Hh signal is not present. Ptc-LacZ Genotype Germarium Later stages Cos2-P2835b Lower Higher Cos2-w1 Lower Higher Figure 4. Levels of Ptc-LacZ in homozygous Cos2 clones compared to wild type 7 and 14 days-post heat shock. This method used Ptc-LacZ stain to approximate Ptc gene expression in response to Cos2 homozygous mutations. 7 days after adult-heat shock, the brightness of Ptc-LacZ stain in clonal cells was compared to that of wild type cells. There are two main limitations in our method of measuring Hh output using Ptc-LacZ. First, we used the expression of one Hh response gene, ptc, as the benchmark to gauge Hh response. But because Ptc is the only one among many others regulated by the Hh signal, we would gain a more complete knowledge had more Hh-regulated genes been known. Nevertheless, Ptc transcription alone should give us a glimpse into the Hh target gene response. Secondly, our approach of evaluating ptc expression using ptc-lacz antibody staining yielded qualitative and approximate results. Briefly, LacZ gene expression was driven by Ptc promoter to produce -gal as product. The brightness of the ptc-lacz antibodies was visually evaluated to approximate Ptc expression. However, the brightness of the stain was affected by artifacts introduced during the stain procedure or antibody quality. Nevertheless, our method warrants a rough approximation of Ptc expression cos2 inactivation temporarily rescued somatic stem cell numbers in Shg mutants. To examine whether cos2 inactivation rescued the loss of stem cells seen in Shg mutants, we generated SSC clones lacking cos2 activity. These Cos-Shg double mutants were negatively marked with endogenous GFP and Fas antibody. We then counted the total number of SSC clones in ovaries in cos2-shg double mutants and compared with that of cos2 or Shg mutants alone 5. Figure 5. The percentages of SSC clones that are homozygous for each of the mutations are measured at 7 and 14 days post-heat shock. Adult flies were heat shocked two times in 37C water bath to induce gene inactivation. Four different fly strains that contained four different mutations, Cos2, shgr69, cos2/shgr69, and cos2/shg10469, were used. 7 and 14 days later homozygous somatic stem cell clones was identified by GFP negative staining and the percentage of ovaries that contained each clone was calculated. Early on, cos2 inactivation resulted in a two to ten-fold increase of somatic stem cells. Specifically, only 3% of clones in shgr69 flies contained miotic somatic stem cells, indicating low proliferative potential. In contrast, in shgr69 and cos2 double mutants, 31% of all clones were from somatic stem cells, marking a dramatic increase SSC s. The cos2 mutation with a different Shg allele yielded the same result-27% of clones in cos2- shg10469 were from somatic stem cells, indicating that the cos2 rescue of Shg phenotypes doesn t depend on the specific Shg allele. Finally, it is noted that shg10469 is a hypomorph and shgr69 is a null. Thus, cos2-shgr69 rescues the loss of somatic stem cells in higher percentages than cos2-shg10469

6 6 THE HEDGEHOG SIGNALING PATHWAY Ci stain Genotype Germarium Late stages Smo 2 ck Normal Normal DCO H2 Slightly high High Smo D16 DCO H2 Normal High Figure 6. The levels of Ci stain in Smo or PKA clones compared to the wild type cells. Fly larva were heat shocked two times in 37C water bath and, 10 and 14 days later, the flies were dissected and stained with anti-ci-155 antibodies and visualized under the confocal microscope. Visual examination was performed to evaluate the levels of Ci antibody brightness in the clones compared to the wild type cells. posssible because shgr69mutants started out with lower stem cell counts than shg Thus it is easier to see the rescue of somatic stem cells in the null mutant shgr69. Similar observations were made for larval heat shocks. 10 days post heat shock, cos2 inactivation also displayed an increase in the percentages of SSCs from 4% (Shg mutation alone) to 18% for and cos2- shg10469 double mutants, respectively Stem cell clones were lost in cos2-null clones. The cos2-stem cells was not maintained overtime, indicating that SSCs are not overproliferating in response to cos2 inactivation. 14 days post adult heat shock, only 6% of cos2-shgr69 and 8% of cos2-shg10469 ovarioles contained SSC clones. These numbers, while still greater than the percentage of somatic stem cell clones in Shg alone (0%), are much lower than that of cos2 alone (12%). Similarly, 14 days post larvae heat shock, only 7% of cos2-shgr69 and 11% of cos2-shg10469 clones were present in somatic stem cells. This indicates that the initial increase in stem cell clones was short-lived. Visual inspection of ovaries revealed that most germaria did not contain any cell clones but some later egg chambers did. Nevertheless, of those that contained clones, none had exclusively cos2-minus cell lineages, indicating that the cos2-null clones never populate the entire the ovariole, a phenotype observed when ptc is mutated. [4] 3.4. PKA inactivation increased Ci levels whereas Smo mutation had no effect. To establish the effects of Smo and PKA mutations on downstream components of the Hh pathway-specifically Ci-we generated mutants homozygous for Smo (alleles D16 and 2) or PKA (alleles B3 and H2) mutations. Ovaries were examined 7 or 10 and 14 days after heat shock-inducible recombination for antibodies against full-length Ci (Ci-155). The levels of Ci-155 were determined based on the brightness of anti-ci antibody stain. The effects of Hh signal hinge on the Ci-mediated regulation of target gene expression. PKA-mutant clones displayed slightly higher Ci levels in the germarium and perceptibly higher levels of Ci in the later stages. See Tab. 6. This observation makes sense considering that Hh is a morphogen that is secreted and concentrated in germarium. Hence, the later egg chambers have little Hh signal, PKA would normally be active and Ci is processed. PKA inactivation renders Ci unable to be phosphorylated for degradation, resulting in higher Ci-155 in egg chambers. The germanium, in contrast, has higher Hh signal, so PKA is more likely to exist in an inactive state. PKA inactivation in the germarium leads to modest effects compared to that in the later stages. This rationale explains for observation that PKAinactivated clones only exhibit slightly higher levels of Ci in the germarium. Our results showed that Smo inacitvation did not affect Ci-155 levels See Fig. 7. We had first expected that, Smo, the transducer of Hh signal which ultimately regulates Ci cleavage, may also affect Ci levels. Our results showed otherwise. Visual examination based on the brightness of Ci-155 antibodies showed that Ci clones in the germarium were no different from wild type in Ci levels See Fig. 7. A similar pattern was shown in the later egg chambers. (Figure 4b) Like ptc-lacz, because we judged the levels of Ci based on the brightness of antibody fluoresce, it is possible that Smo inactivation altered the levels of full-length Ci. The difference was not large enough to register a perceptible change visually. Similarly, our observation that Smo clones exhibited normal Ci in the later stages also makes sense. Since Smo is a transducer of Hh signal, its inactivation would have small effects if there is little Hh in egg chambers. However, the interpretation of PKA-smo double or Smo mutants in the germarium is less straightforward because Ci may also be processed through a different, PKA-independent ubiquitin complex when Hh is high. It is well accepted that, without Hh signal, Ci- 155 is processed into Ci-75. But, as Ci studies in the wing discs show, when Hh signal is present, there are two modes of regulation. First, the PKA-dependent pathway is activated when low levels of Hh signal ultimately inhibits PKA, freeing Ci-155 from being processed into Ci-75. However, there is also an oppsosite second level of regulation: high levels of Hh prompt the complete degradation of Ci-155 through a PKA-independent ubiqiuin complex. The rates of proteolytic cleavage are different and it is not known which pathway has a greater effect on Ci processing. Furthermore, both the PKA-dependent and PKA-independent degradation may be at work in Smo mutants in the germarium. It is reasonable to assume that Hh signals are higher in the germarium because of the proximity to Hh-secreting cells. We first showed that, in the germaium, Smo mutation alone does not lead to any noticebale differences in Ci levels. Smo is the first transducer of Hh signal and its activation may be indicative of the Hh signal. This observation further supported our speculation that the regulation of Ci processing may not be entirely Hh-dependent. Finally, like cos2, our method of evaluating Ci-155 levels is qualitative and may introduce artifacts during the staining process. Furthermore, Ci processing is very likely to be under the dual control

7 REFERENCES 7 Figure 7. Levels of anti-ci-155 antibody stain. Fly larva were heat shocked two times in 37C water bath and, 10 and 14 days later, the flies were dissected and stained with anti-ci-155 antibodies and visualized under the confocal microscope. The white arrows indicate the location of the homozygous mutant clones. Smo2ck clones were shown in A; DCOH2 clones were shown in B; SmoD16DCOH2 double mutant clones were shown in C. of PKA-dependent and PKA-independent ubiquitination. These two pathways are known to have opposing effects but with unknown magnitudes of influence. 4. Discussion Recently, the theory that certain tumors are initiated from cancer stem cells is strengthened by the finding that these cells are capable of self-renewal and differentiation, characteristics of stem cells. [8] It is important to catalogue alterations in the regulators of stem cell behaviors to understand the mechanisms behind tumorgenesis. One of the first factors that directly regulate stem cell proliferation, Hedgehog may provide crucial clues to stem cell-initiated transformation. This paper uses mutant analysis to approach this question. Most of our observations may be explained considering that Hh plays a multifaceted role in multiple cellular processes. Specifically, we found that both cos2-inactivated stem cells and their progeny follicle cells exhibited increased ptc expression. It is not completely surprising that cos2 inactivation rescued the loss-of-stemcell phenotype in Shg mutations considering that Hh signal regulates many target genes. cos2 inactivation may affect profiles of adhesion proteins by altering Hh signal. For example, the Hh response genes may code for adhesion proteins that were upregulated in cos2-inactivated cells. This would mediate the adverse effects of the Shg mutation. However, this proposal would need to be tested in future studies that probe into the molecular mechanism behind the cos2 rescue. Furthermore, our results showed that the cos2-inactivated stem cell clones are lost overtime. While the molecular details remain unclear, our observation suggests that cos2 inactivation did not increase the stem cells ability to divide. We cannot confirm the fate of the cos2-minus stem cells because it is not yet possible to track them overtime. Nevertheless, we speculate that they may have either moved out of the stem cell niche or died. First, it is thought that there are about three positions in a germarium that contain the correct combination of extracellular signals for proper somatic stem cell maintenance. [4] Since Hh is a key environmental cue for somatic stem cell behavior, alterations in the pathway may cause the clones to leave the stem cell niche and differentiate into pre-follicle cells. This scenario would be consistent with our observation that cos2-inactivation exists as follicle cells in egg chambers but not as somatic stem cells in the germarium-the stem cell clone initially divides asymmetrically to generate a lineage of clones before differentiating into one itself. Alternatively, mutations in cos2 may render the clones unable to survive for long. A time-course tracking experiment would confirm the ultimate fate of these somatic cell clones. Our results provide further evidence that alterations in Hh alter stem cell behaviors by affecting target gene expression. A number of questions still need to be elucidated before we can conclude whether Hh is implicated in stem cell-initiated tumorigenesis. Future studies that can selectively label Hh mutation clones may lend insights into the molecular consequences of Hh alterations on stem cell behaviors. Nevertheless, it will be exciting to envision the clinical development of inhibitors against Hh in cancer therapeutics. References [1] Rubin LL, de Sauvage FJ. Targeting the Hedgehog pathway in cancer. Nat Rev Drug Discov Dec;5(12): Review. [2] Kirilly D, Xie T. The Drosophila ovary: an active stem cell community. Cell Res Jan;17(1): Review. Erratum in: Cell Res Mar;17(3):271.

8 8 THE HEDGEHOG SIGNALING PATHWAY [3] Faurschou A, Haedersdal M, Poulsen T, Wulf HC. Squamous cell carcinoma induced by ultraviolet radiation originates from cells of the hair follicle in mice. Exp Dermatol Jun;16(6): [4] Zhang Y, Kalderon D. Hedgehog acts as a somatic stem cell factor in the Drosophila ovary Nature 2001 Mar; 29 (410): This Letter to Nature article shows that hedgehog acts directly on the proliferation and differentiation of stem cells in the Drosophila ovary. [5] Harrison DA, Perrimon N Simple and efficient generation of marked clones in Drosophila. Curr Biol Jul 1;3(7): [6] Forbes, A.J., Spradling, A.C., Ingham, P.W., Lin, H. The role of segment polarity genes during early oogenesis in Drosophila. Development 122, (1996). [7] Beachy, P.A., Karhadkar, Berman, Tissue repair and stem cell renewal in carcinogenesis. Nature Nov 18;432(7015): [8] Athar M, Tang X, Lee JL, Kopelovich L, Kim A Hedgehog signalling in skin development and cancer. Exp Dermatol Sep;15(9): [9] Zhang, Y. Kalderon d. Regulation of cell proliferation and patterning in Drosophila oogenesis by Hedgehog signaling, Development : [10] Ho KS, Suyama K, Fish M, Scott MP. Differential regulation of Hedgehog target gene transcription by Costal2 and Suppressor offused. Development Mar;132(6):

9 Personality Perception in Toddlers: Trait Awareness in Four-Year-Olds as Related to Threat, Consistency, and Behavior A Prospectus for an Honors Thesis in the Department of Psychology Margaret Arnold 1 1. Introduction Throughout the course of everyday life, we are constantly encountering new people and the personalities they present to the world. While we would like to be able to process and remember every facet of every person s personality, this is an incredibly difficult feat. Research shows that not all traits are weighted equally, and certain types of traits are privileged over others in memory and judgment. The current work explores this concept by examining the ways in which 4-year olds perceive their peers. The results of this research will help us understand the types of information young children pay attention to in others, and how they may go about categorizing and remembering this information. I hypothesize that there will be a number of seemingly different traits that 4-year olds will remember with accuracy. There may be, however, over-arching themes that determine young children s choices in judging others. Such themes include social relevance, level of threat, group membership, consistency, valence, and extremity 2. Because there is a limited amount of information one can retain, children will scale the importance of traits based on such themes. The traits that do not fall into these categories will not be remembered as accurately. The results of this study will thus contribute immensely to the understanding of world view and personality perception in 4-year olds. My main question in this research is: Why do toddlers remember the things they do about their peers? Why are certain things more important than others, and what does this tell us about the human psyche? If toddlers do in fact judge traits based on threat and social relevance, how do they decide what things are relevant to them in the first place? I will conduct a brief review of the research on personality in order to provide a foundation for the current study and discuss why the understanding of trait perception is important. 2. Brief Review of Adult Personality Research Research in adult personality and social psychology shows that there are specific categories of traits that may be remembered better than others. Of particular relevance are studies examining the importance of valence and extremity of traits, and how these categories may contribute to accurate recall. Fiske (1980) found in a study of adults that extreme, negative behaviors receive more attention than both moderate, negative behaviors and positive behaviors, regardless of extremity. In this study, subjects viewed two pre-scaled behavior photographs for each of 17 stimulus faces, and looking time was recorded for each face. Results showed that looking time was longest for negative, extreme behaviors across the board. People tended to respond most sensitively to behaviors that fell out of the ordinary, a concept of significant relevance to the present study. Various other studies have confirmed the tendency for adults to rely on the extremity of traits when making judgments of others. Rosenberg & Olshan (1970) found that adults will make evaluative judgments of others by situating traits at two extremes. For example, traits like bold and timid are seen as polar opposites, and adults will judge others based on this distinction. In a study of the Five Factor Model of personality 3, McLarney et al (2006) found the highest rater agreement on extraversion in judging others on the Big Five traits. This indicates that extraversion is a highly salient trait, and that people tend to agree on the types of things that label a person an extrovert, as opposed to an introvert. There have been significant findings in adult research indicating the importance of negative information in memory for faces and for words. Kensinger & Corkin (2003) found that adults are more likely to remember negative information than neutral information. 1 Margaret Arnold 08 is a psychology concentrator and 2007 PRISE Fellow. 2 A definition of some basic concepts is in order. Consistency is defined by the reliability or uniformity of successive results or events. Valence is defined as the degree of attraction or aversion than an individual feels towards a specific object or event. Extremity refers to the character or kind farthest removed from the ordinary or average. 3 The Big five personality traits are five broad factors or dimension of personality discovered through empirical research. They can be summarized as follows: Openness: active imagination, aesthetic sensitivity, attentiveness to inner feelings, preference for variety, intellectual curiosity Conscientiousness: a tendency to show self-discipline, act dutifully, and aim for achievement; planned rather than spontaneous behavior Extraversion: energy, positive emotions, and the tendency to seek stimulation and the company of others. Agreeableness: a tendency to be compassionate and cooperative rather than suspicious and antagonistic towards others. Neuroticism: a tendency to experience unpleasant emotions easily, such as anger, anxiety, depression, or vulnerability; sometimes called emotional instability. 9

10 10 PERSONALITY PERCEPTION IN TODDLERS Negative information was also recalled with greater, more vivid detail than neutral information. Further research suggests that faces paired with negative information are more likely to be remembered as well. In a study of college students, Mealey, Daood, & Krage (1996) presented subjects with photographs of faces paired with two pieces of information. The first piece of information was the photographed person s social status (high or low), and the second was the person s character (history of cheating, irrelevant information, or history of trustworthiness). A week later, subjects were again presented with the photographs and asked which ones they thought they had seen before. Results show that subjects had the highest recall for the faces of cheaters. As mentioned above, cheating would fall into the category of an evaluative trait, indicating a given person s morality, character, and threat potential. Adults also seem to make distinctions between traits related to morality and traits related to ability. Martijn, Spears, Van der Pligt, & Jakobs (1992), for example, found that negative behavior is informative in making judgments about morals, but positive information is more important in making judgments about ability. These themes of valence and extremity, particularly when paired with research on visual memory, will certainly be important in understanding personality perception in 4-year olds. 3. Brief Review of Personality Research in Toddlers This background adult research paves the way for a number of recent studies done in toddlers. This available research in personality perception may be helpful in explaining why certain traits are more important than others to toddlers. Heyman & Legare (2005) investigated children s assessment of the value of various sources of trait information in a study of 6- to 7-year-olds and 10- to 11-year-olds. This, as well as other research has shown that children are likely to assume that personality traits in others are consistent. For example, if a young boy picks up on one negative trait in another child, he will assume that this same child is characterized by other negative traits as well (Heyman & Legare, 2005). Rholes & Ruble (1984) found a similar effect in a study of children s understandings of dispositional characteristics. In this study, participants observed vignettes meant to reveal certain characteristics about the actor s personality. Participants were then asked to make predictions about what the actor was going to do next. Results show that children expect others to be consistent, relying heavily on overt behavior to make inferences about personality. In a separate study, Klein et al (1992) found that knowledge of a person s personality is often largely based on memory of that person s past behavior. This will be particularly relevant to the current study, as participants are presented with faces accompanied by trait labels and associated overt behaviors. Not only do children assume that others behavior is consistent, but they are also display aversion to inconsistent behavior. In a study of 3- to 5-year-olds, Giles & Heyman (2005) found that children were more likely to approve of aggressive behavior against a person labeled as mean than against a person labeled as nice. They were not against hitting a person who had been labeled as mean and who had done something inconsiderate to another. These studies thus demonstrate two main things about the ways in which toddlers judge their peers. First, toddlers distinguish traits based on how positive or negative they are. Second, toddlers prefer that a person s behavior be consistent. They predict people s future behavior and overall personality based on their behavior in the present. There are, however, situations in which people are not characterized by black-and-white labels as nice and mean. How do toddlers go beyond these basic traits to decide what information is important? Heyman & Gelman (1999) showed that children will assign dispositional labels, like shy, to people in order to predict future behaviors. It seems that children perhaps assign valence to certain traits according to how relevant they are and what they may indicate about a person s intentions. As mentioned previously, Heyman & Legare (2005) conducted a similar study on 6-7 year-old and year-old subjects. The purpose of this study was to show that children remember behaviors differently based on whether the traits involved are evaluative or non-evaluative. 4 Evaluative traits (i.e. smart, honest ) are those that indicate something about a person s intelligence or morality. Non-evaluative traits comparison (i.e. outgoing, nervous ) are those that indicate something about a person s temperament or ability. Subjects were asked a pair of questions about each of 4 traits: smart, honest, outgoing, and nervous. 5 An example of one such question is, If you told someone that you just met that you thought they were very honest, how do you think they would feel? The goal was to determine what types of traits (evaluative or non-evaluative) kids care about most. Results show that kids have a higher sensitivity to the valence of evaluative traits (honest, smart) than non-evaluative traits (outgoing, nervous). Thus, when questions were asked about extreme traits, kids were most concerned about the extreme evaluative traits. This brief review highlights the possible explanations for why 4-year olds may privilege certain traits over others. Social relevance, level of threat, valence, extremity, group membership, and consistency are the overarching themes toddlers may rely on to place information into distinct categories. Recent studies of face memory suggest that toddlers are good at remembering information related to threat (Kinzler & Shutts, in press). In Experiment 1, Kinzler & Shutts (in press) presented 3 and 4-year olds subjects with a series of 8 adult faces. Each face was accompanied by a trait label ( nice or mean ) and a behavior consistent with this trait. After familiarization, subjects were presented with the 8 adult faces again, but each was accompanied by a distracter face. Subjects were then asked which face they had seen before. Results of Experiment 1 show that memory was higher for individuals who had done mean things than for individuals who had done nice things. In Experiment 2, Kinzler & Shutts ((in press) presented 3 and 4-year old subjects with a series of 8 adult faces. Each face was described as having received either a mean action or a nice action. Subjects did not show a memory bias for individuals 4 In these studies, evaluative traits are defined as traits that vary along a single dimension in which one value is clearly positive and another is clearly negative. Hence for toddlers evaluative traits can be thought of as judgments that refer to the inherent goodness or badness of an actor. 5 In these studies, the traits were defined as follows: Smart people are good at learning new things. Honest people tell the truth and keep their promises Outgoing people like to be around people and get attention from people Nervous people worry a lot about things.

11 4. PILOT STUDY 11 who had experienced a mean action over individuals who had experienced a nice action. These combined results suggest that the memory bias seen in Experiment 1 was due to threat, and not negative valence, of the trait. These intriguing results serve as the springboard for the current project. The remaining question is why do children remember these traits? The work of Kinzler & Shutts (in press) proposes two explanations for subjects performance in the study. First, toddlers may simply be experiencing feelings of general arousal when presented with threatening information. Their brains are stimulated by this sense of threat and it is thus stored in memory. Another possibility is that toddlers actually place various types of information into different categories. Toddlers decide that certain categories are more important than others to remember. Subjects in Kinzler & Shutts ((in press) study have thus decided that threatening information is worth remembering, and they distinguish between individuals based on this category of information. There may also be a series of additional factors involved other than threat of the trait. 4. Pilot Study As previously described, studies have demonstrated two main things about the ways in which toddlers judge their peers. First, toddlers distinguish traits based on how positive or negative they are. Second, toddlers prefer that a person s behavior be consistent. The current study will address these phenomena, and expand on the complexity of trait perception to answer the more difficult questions of why toddlers pay attention to these things, and why toddlers make these assumptions. The study consists of a number of conditions examining different personality traits in order to determine the types of information 4-year olds attend to privilege in memory. Four-year-old participants view a brief slide show in which they are introduced to eight children. For each face presented, participants are given a trait label and an overt behavior consistent with this trait. Participants are then shown the faces again and asked to remember the trait label they had been told. In additional conditions, participants will be asked what specific overt behaviors they can remember. I predict that 4-year olds will choose which traits to store in memory according to social relevance, level of threat, group membership, and consistency. Four-year olds will not remember ambiguous traits that may seem irrelevant or unimportant. This study will be a significant step in reconciling past research on personality perception in adults and toddlers. The current project will thus build on past work and closely examine the variety of factors at work behind personality perception in 4-year olds. I hope to come up with a more concrete explanation for why certain traits are remembered better than others Experiment 1: Memory Bias for the Nice/Mean vs. Inside/Outside condition in adult faces. Purpose. The purpose of Experiment 1 was to extend the research of Kinzler & Shutts (in prep) and determine why children attend to this trait information in adult faces. The Nice/Mean condition was compared to a control condition, Inside/Outside, to begin to explore the types of traits that are important to 4-year old subjects. Methods Participants. Participants in the current study were 4-year old children recruited from the Harvard Lab for Developmental Studies database. Subjects were predominantly white and lived in the Boston area. A total of 48 subjects (24 per condition; 28 males, 20 females) participated in the study. Upon completing the study, participants received a Harvard Lab for Developmental Studies stuffed animal, t-shirt, or toy, and the parent received 5 dollars reimbursement for travel. The experimenter also debriefed parents on the study and allowed them to ask any questions. Materials and Procedure. Experiment 1 was conducted in the format of a brief slide show on a laptop computer, and took a total of about eight minutes. Participants were seated in front of the laptop and told to pay attention to the slide show on the screen. The experimenter went through a series of eight adults faces matched for attractiveness, and introduced each one to the participant along with a trait label and an accompanying behavior (e.g. This is Jessica. She is always nice. Today she brought in cookies and everyone got some ). The experimenter then went through the eight adults faces again in a different order and asked the participant if he or she could remember that original trait label (e.g. Do you remember, was Jessica nice or mean? ). This condition was thus called the Nice/Mean condition. The control condition (Inside/Outside) looked at a more ambiguous piece of information: playing inside versus playing outside. The Inside/Outside condition used the same format as the Nice/Mean condition. Participants were once again introduced to the adult faces with a trait label and an accompanying behavior (e.g. This is Kimberly. She always plays outside. Today she threw a ball with some kids ), and then asked whether or not they could remember that label (e.g. Do you remember, was Kimberly playing inside our outside? ). This control condition (Inside/Outside) served as a means of comparison to the experimental condition (Nice/Mean). Design and Analysis. All conditions were counterbalanced to account for ordering effects. There were a total of four power-point slide shows: two beginning with male faces, two beginning with female faces. Within each of these four possible slide shows, there were two possible scripts. One script began by describing the child as mean (or playing outside in the control condition) and the other began by describing the child as nice (or playing inside in the control condition). There were thus eight possible scripts for each condition, and four possible slide shows. General means for each condition were taken to examine participants accuracy on the task. Results within each condition were then analyzed using one sample t-tests. Results between conditions were analyzed using independent sample t-tests. Results Participants performed above chance in remembering the trait labels nice and mean in the experimental condition (Chance = 50%; M = 64.06%; SE=.04; t(23)= 3.288, p.005). Participants in the Inside/Outside condition performed at chance (Chance = 50%; M = 51.56%; SE=.036; t(23)=.430, p= n.s.). The Nice/Mean condition was also found to be significantly different from the Inside/Outside condition (M = 64.06%; SE=.04; t(46)= 2.228, p.05). Discussion These results suggest that 4-year olds show a memory bias for the trait labels nice and mean in adult faces. They do not, however, remember traits that are seemingly unimportant and trivial. Experiment 1 is limited, however, because the

12 12 PERSONALITY PERCEPTION IN TODDLERS stimuli used are adult faces. Would subjects show the same effect if they were introduced to children accompanied by similar trait labels? Experiment 2 explores this possibility, as well as an additional condition looking at different traits Experiment 2: Memory Bias for Nice/Mean, Introvert/Extrovert and Inside/Outside in Children s faces. Purpose. While the results of Experiment 1 are intriguing, this study is only one step along the way to forming a complete picture of peer perception in 4-year olds. Experiment 2 goes beyond Experiment 1 in two ways. First, it uses children s faces as opposed to adult faces. This study thus questions whether toddlers remember the same categories of information in children (their peers) as they do in adults (their authorities). Second, Experiment 2 includes a third condition (Introvert/Extrovert) examining the traits talks a lot and quiet. The Introvert/Extrovert and Nice/Mean conditions are similar in that they both examine socially relevant traits. The traits in these two conditions are indicators of an individuals overall temperament and may also place individuals into social groups. The Introvert/Extrovert condition is unique from the Nice/Mean condition, however, because it is not necessarily associated with threat. Whether a person is quiet or talking a lot may be considered a trait that will be consistent over time. The Introvert/Extrovert condition thus suggests that it is not just a trait s threat value, but also its social relevance and consistency, that makes it important for toddlers to remember. Methods Participants. Participants for Experiment 2 were recruited from the same database used in Experiment 1. A total of 96 4-year olds (32 per condition; 39 males, 57 females) completed the study. As in Experiment 1, subjects were predominantly white and lived in the Boston area. A portion of the subjects were recruited from a school in Lincoln, Massachusetts, a suburban, predominantly white town. Materials and Procedure. Experiment 2 used the same materials and procedure used in Experiment 1 expect for the following exceptions: stimuli were children s faces rather than adult faces, and there were three conditions (Nice/Mean, Inside/Outside, Introvert/Extrovert) as opposed to two. In the Introvert/Extrovert, subjects were introduced to children labeled talks a lot or quiet (e.g. This is Drew. He is always quiet. Today he made a painting in the playroom ). Subjects were than asked whether or not they could remember this trait (e.g. Do you remember, was Drew quiet or talking a lot? ). Data and Analysis. The same methods used in Experiment 1 were also used in Experiment 2 for counterbalancing and analyzing data. The Introvert/Extrovert condition allowed for a third point of comparison to the control condition through independent sample t-tests. Results As in Experiment 1, participants in Experiment 2 performed above chance in the Nice/Mean condition (Chance = 50%; M = 66.02%; SE=.0338; t(31)= , p.001). Performance on the Nice/Mean condition was also significantly different than the Inside/Outside control condition (M = 66.02%; SE =.034; t(62)= 2.405, p.01). Of particular interest in Experiment 2 is that participants performed above chance in the Introvert/Extrovert condition (Chance = 50%; M = 65.23%; SE=.034; t (31)= 4.478, p.001). Additionally, this condition was significantly different than the Inside/Outside control condition (M = 65.23%; SE =.034; t(62)= 2.212, p.05). Also important to note are the results of the Inside/Outside condition in Experiment 2 as compared to those of Experiment 1. As mentioned previously, subjects in Experiment 1 performed at chance in the Inside/Outside condition (Chance = 50%; M = 51.56%; SE=.036; t(23)=.430, p= n.s.), significantly worse than they did in the Nice/Mean condition. Subjects in Experiment 2 showed a similar effect, performing close to chance in the Inside/Outside condition (Chance = 50%; M = 55.73%; SE =.026; t(31)= 2.183, p.05) and significantly worse than in both the Nice/Mean condition and the Introvert/Extrovert condition. While the Inside/Outside conditions in Experiment 1 and 2 are not significantly different from each other, it is important to note that subjects in Experiment 2 performed above chance. Subjects in the Inside/Outside condition of Experiment 1, however, performed at chance. It appears that perhaps 4-year olds attend more closely to traits labels for children s faces (Experiment 2) than adults faces (Experiment 1). Another possibility is that 4-year olds have trouble imagining an adult playing on a playground or reading a book in the classroom. Thus, social relevance may be an explanation for this category of trait judgment. The results could also simply be a product of the stimuli used in the two experiments. Perhaps subjects had an easier time distinguishing between the children s faces in Experiment 2 than the adults faces in Experiment 1. See Fig. 1. Figure 1. Results for the three conditions in Experiment 2 (children s faces).

13 5. GENERAL DISCUSSION 13 Discussion The results of Experiment 2 are promising for a number of reasons. First, they give direct insight into 4-year olds peer perception because the stimuli used are children s faces. Given that peers can potentially have a large impact on a young child s life, it is interesting to note the ways in which 4-year olds may judge and evaluate them. A comparison of the results of Experiment 1 and Experiment 2 may also serve to explain the value of the social relevance of traits in toddlers personality perception. Second, Experiment 2 not only replicates Experiment 1 (by showing that toddlers perform above chance in the Nice/Mean condition), but it also shows that toddlers perform above chance in the third condition (Introvert/Extrovert). Experiment 2, particularly the results of the Introvert/Extrovert condition, adds depth to the current work, but it also presents a number of lingering questions. In order to understand these results, we must examine how the two experimental conditions are related. The hypothesis of Experiment 1 was that toddlers would remember information indicating level of threat, thus causing them to remember the traits nice and mean. In Experiment 2, the Introvert/Extrovert condition suggests that toddlers are remembering information based on social relevance and consistency of the trait. Perhaps subjects in the Introvert/Extrovert condition are using these traits as judgments of personality and placing children into social groups based on this. Past research also shows that children value consistency in perceiving other. Perhaps subjects consider the labels quiet and talking a lot to be measures of temperament that will help them predict future behavior. Subjects would thus store this information and perform well in the Introvert/Extrovert condition. The focus of this discussion is the question of how exactly 4-year olds choose to categorize information, and how various types of traits may be related to one another. Further research is necessary to explore this question further, and to understand the minute details behind trait perception in toddlers Experiment 3: Forced-Choice Condition in Behaviors Related to Threat. Purpose. Experiments 1 and 2 have provided insight into 4-year olds personality perception. Past research and the results of these two studies have suggested that 4-year olds attend to traits in others for a number of possible reasons. Of particular interest are the explanations that toddlers attend to traits indicating level of threat, social relevance, and consistency. Kinzler & Shutts in press suggest that toddlers privilege threatening information ( mean ) over non-threatening information ( nice ). While subjects in Experiments 1 and 2 performed above chance in the Nice/Mean condition, there was no memory bias found for specifically mean information over nice information. The purpose of Experiment 3 will thus be to go deeper into these quandaries by using a Forced Choice condition. Will this more sensitive measure cause toddlers to show a memory bias for mean actions over nice actions? In addition, will subjects remember more detail about the specific behaviors associated with being mean as opposed to being nice? Results of this study will show whether toddlers have a higher preference for labels and overt behaviors directly related to threat. The Forced Choice format will be momentous in making connections between past research and the current project. Methods Participants. Participants in this study will be drawn from the same database used for Experiments 1 and 2. If the results of the first 8 subjects show a strong effect, the study will include a total of approximately 30 predominantly white subjects. Materials and Procedure. The method for Experiment 3 will be identical to that of Experiment 1 expect for a few exceptions. The scripts will be adjusted slightly to make this a Forced Choice condition. Subjects will be introduced to 8 children s faces in a brief slide show using the same format as Experiment 1. In Trial Phase 1, subjects will be forced to choose between either two mean actions or two nice actions (i.e. Did Briana push someone on the playground or yell at kids and make them cry? ). After Trial Phase 1 is complete, subjects will be introduced to the 8 children s faces for a second time. In Trial Phase 2, subjects will be forced to choose between a mean action and a nice action (i.e. Did Christopher bring in cookies and everyone got some or steal everyone s cookies and no one got any? ). Data and Analysis. The results of Experiment 3 will be analyzed using general means and one sample t-tests. If a control condition is conducted, independent sample t-tests will be used to compare the results of both conditions. Experiment 3 will also use the same counterbalancing techniques used in the two previous studies to account for ordering effects. Expected Results The results of this study will be informative in untangling the mechanisms behind 4-year olds perceptions of threat in their peers. I expect that subjects will be good at remembering overt mean behaviors. Thus, they will perform well above chance in Test Phase 2. Subjects will also be able to give more detailed responses for the mean behaviors than for the nice behaviors. I do not expect that subjects will perform incredibly well in Test Phase 1, when they are asked to distinguish between either two mean actions or two nice actions. I think that toddlers tune in closely on the general idea of a person behaving mean actions, but they do not necessarily distinguish between two different mean actions. Subjects will remember that a child was mean, but they will not always know what behavior he or she displayed in the familiarization phase. If there are significant effects in Experiment 3, I will continue this Forced Choice condition using other traits (Inside/Outside, Introvert/Extrovert). Depending on these results, I may also run conditions examining physical attributes, such as Tall/Short or Fat/Thin. It would be interested to see whether 4-year olds may a distinction between internal personality traits and external physical traits that a person cannot control. The combination of all of this research will present a fairly thorough picture of personality perception in 4-year olds. 5. General Discussion The ability to perceive others is a vital aspect of survival. Because we need to be able to make snap judgments about the environment, our information-processing system is forced to create an imbalance between certain types of information. While researchers have made significant efforts to model the ways in which adults categorize and conceptualize others, there is a need for more empirical study in toddlers. The current research is an effort to elaborate on the available research in

14 14 PERSONALITY PERCEPTION IN TODDLERS toddlers, and make strides towards a full understanding of the information 4-year-olds use to interpret their peers. This project will have a number of implications for the ways in which young toddlers evaluate those around them. I hope to answer why toddlers remember certain traits more than others, and what types of information toddlers use to make the decision to remember a given trait. The results of Experiments 1 and 2 are promising, and Experiment 3 should being the trends found thus far to a deeper level. The roles of social relevance, level of threat, group membership, consistency, valence, and extremity of the trait will all come into play throughout the project. By comparing the results of each condition, I will be able to understand the categories of information toddlers may use to distinguish between certain traits. If social relevance is the primary category used by toddlers in judging their peers, there may be significant differences in the things children in comparison to adults remember. Four-year olds may also remember different types of information when they are paired with an adult s face as opposed to a child s face. If other categories, such as valence or extremity, are significant forces subjects memory for trait labels, 4-year olds may be more similar to adults than one would think. In completing this study, I hope to gain a thorough understanding of the ways in which toddlers judge others, as well as the methods they use to decide what is worth remembering. This project also raises a number of lingering questions in the field of developmental psychology. When exactly does this ability to perceive others develop? Research suggests that the natural instinct to categorize information seems to grow with age, and young children begin to formulate inner schemas for evaluating peers and predicting future behavior. The roots of peer perception, however, may actually lie in infancy. In a study of facial recognition in infants, Meltzoff (1995) states that newborns begin life with a grasp of people. Infants seem to place much of their attention on faces associated with fear and anger (Montague & Walker-Andrews, 2001). Thus, in order to fully understand personality perception in toddlers, we may have to begin with an exploration of the available research in infants. One s ability to make judgments and store information begins at an early age, and the progression to adulthood can be seen as a gradual, continuous process over time. References [1] Fiske, S. T. (1980). Attention and weight in person perception: the impact of negative and extreme behavior. Journal of Personality and Social Psychology, 38, [2] Giles, J. & Heyman, G. D. (2005). Preschoolers use trait-relevant information to evaluate the appropriateness of an aggressive response. Aggressive Behavior, 31, [3] Heyman, G.D. (1999). The use of trait labels in making psychological inferences. Child Development, 70, [4] Heyman, G. D. (2001). Children s interpretation of ambiguous behavior: evidence of a boys are bad bias. Social Development, 10, [5] Heyman, G. D. & Giles, J. (2004). Valence effects in reasoning about evaluative traits. Merrill-Palmer Quarterly, 50, [6] Heyman, G. D. & Legare, C. H. (2005). Children s evaluation of sources of information about traits. Developmental Psychology, 41, [7] Kensinger, E. A. & Corkin, S. (2003). Memory enhancement for emotional words: Are emotional words more vividly remembered than neutral words? Memory & Cognition, 31, [8] Kinzler, K. & Shutts, K. (in prep). The influence of threat on children s face memory. Klein, S., Loftus, J., Trafton, J. G., & Fuhrman, R. W. (1992). Use of exemplars and abstractions in trait judgments: a model of trait knowledge about the self and others. Journal of Personality and Social Psychology, 63, [9] Martijn, C., Spears, R., Van der Pligt, J., & Jakobs, E. (1992). Negativity and positivity effects in person perception and inference: Ability versus morality. European Journal of Social Psychology, 22, [10] McLarney-Vesotski, A. R., Bernieri, F., & Rempala, D. (2006). Personality perception: a developmental study. Journal of Research in Personality, 40, [11] Mealey, L, Daood, C., Krage, M. (1996). Enhanced memory for faces of cheaters. Ethology & Sociobiology, 17, [12] Meltzoff, A. N. (1995). Infants understanding of people and things: From body imitation to folk psychology. In J. L. Bermdex, A. Marcel, & N. Elian (Eds.), The body and the self (pp ). Cambridge, MA: MIT. [13] Montague, D. P. F., & Walker-Andrews, A. S. (1997). Peekaboo: A new look at infants perception of emotion expressions. Developmental Psychology, 37, [14] Rholes, W. & Ruble, D. N. (1984) Children s understanding of dispositional characteristics of others. Child Development, 55, [15] Rosenberg, S. & Olshan, K. (1970). Evaluative and descriptive aspects in personality perception. Journal of Personality and Social Psychology, 16, Appendix 1. Nice/Mean Condition Script In this game we re gong to meet some people and learn some things about them. This is Clifford - he s a big red dog. This is blue - he s a big blue dog. Is Clifford a big red dog or a big blue dog? Is blue a big blue dog or a big red dog? Now we re going to see Briana. Briana is always Mean - today she pushed someone on the playground Now we re going to see Jessica Jessica is always Nice - today she helped someone on the playground. Was Brianna nice or mean? Was Jessica nice or mean? 2. Inside/Outside Condition Script In this game we re gong to meet some people and learn some things about them. Now we re going to see Jessica Jessica is always playing Inside - today she baked cookies with her friends. Now we re going to see Briana. Briana is always playing Outside - today she played in the sandbox with her friends. Did Jessica play Inside or Outside? Did Briana play Inside or Outside?

15 4. NICE/MEAN FORCED CHOICE CONDITION SCRIPT Introvert/Extrovert Condition Script In this game we re gong to meet some people and learn some things about them. Now we re going to see Briana. Briana always talks a lot - today she cheered during the kickball game after school. Now we re going to see Jessica Jessica is always quiet - today she watched a movie after school. Was Jessica quiet or talking a lot? Was Briana quiet or talking a lot? 4. Nice/Mean Forced Choice Condition Script In this game we re gong to meet some people and learn some things about them. Now we re going to see Briana. Briana is always Mean - today she pushed someone on the playground Now we re going to see Jessica Jessica is always Nice - today she helped someone on the playground. 1. Did Briana push someone on the playground or yell at kids and make them cry? 2. Did Jessica help someone on the playground or say nice things to kids and make them smile?

16 Reason and the Passions: From Plato to Neuroeconomics Dmitry Taubinsky 1 Reason is ought only to be the slave of the passions, and can pretend to no other office, but to serve and obey them. (Hume, Treatise of Human Nature 2.3.3). 1. Introduction We are often tempted to believe that something of our identity, something deep in our human nature, survives the temporal constraints of our bodies and reaches out beyond the finitude of our world. We call this our soul and spend our lives wondering what that really means (until, perhaps, we have a fling with existentialism). The temptation is to find something immutable, deeper than our particular actions, desires, or feelings, something about ourselves that stays fixed. For Plato, while the soul was immortal, its nature was never fixed but constantly marked by strife: Her [the soul s] form may be described in a figure as a composite nature made up of a charioteer and a pair of winged steeds, the one a noble animal who is guided by word and admonition only, the other an ill-looking villain who will hardly yield to blow or spur. Together all three, who are a figure of the soul, approach the vision of love. And now a fierce conflict begins and at last the charioteer, throwing himself backwards, forces the bit out of the clenched teeth of the brute, and pulling harder than ever at the reins, covers his tongue and jaws with blood, and forces him to rest his legs and haunches with pain upon the ground. (Plato, Phaedrus) So for Plato, the soul, or the human psyche, was not a unity but composed of multiple parts. On the one hand there was the charioteer-cool, calculating, with a clear destination in mind. Yet, by himself, the charioteer could not get to his destinationto do so, he needed to harness the power of his steeds. Thus, the farsighted, goal-directed charioteer was endowed with the somewhat unruly, sometimes aimless, but tremendously powerful steeds. The chariot was propelled by two different forces: the passionate, but directionless steeds on the one hand, and the calm, controlling hand of the charioteer on the other. I summarize Plato s model of the psyche as consisting of two components: 1. the affective system, which is the source of desire and immediate motivation and 2. the deliberative system, which is much more farsighted than the first, and thus keeps the first in check to make sure that its immediate desires do not sabotage the long-term goals. 2 When do we see these two systems in conflict? We see them in conflict when the smell of pie awakens our appetite, while the charioteer in the back of our minds reminds us of how hard it will be to work off this pie at the gym. Or when a particularly obnoxious friend causes us to grind our teeth and put up our fists, while something in the back of our minds quietly tells us that it may be more prudent to keep calm. Or when the sound of falling money in a casino makes us certain that this is our special day, while at the same time reason tells us that our chances of winning are no better than those of any other sucker feeding the casino. Thus a model of the kind that Plato imagined, which allows for multiple sources of motivation to conflict within the human psyche by pitting reason against the passions, best conforms to human behavior than a vision of a unitary, singly-directed soul. Traditional economics, however, assumes the latter. In section 1 of this essay, I will discuss the assumptions made by the rational-agent model in traditional economics. In particular, I will discuss 1.1) Utility maximization and self-regarding preferences 1.2) Expected utility theory 1.3) Stable preferences over time and 1.4) money as a unit of exchange. In section 2, I will show that human behavior contradicts the traditional assumptions in 1.1, 1.2,1.3, and 1.4. In section 3, I will discuss the neuroscience data that gives a deeper grounding to these behavioral anomalies. I will conclude by describing the two-system model of human behavior that emerges as a result of the recent neuroeconomic findings discussed in Section 3. 1 Dmitry Taubinsky 09 is an economics concentrator and 2007 PRISE Fellow. 2 The two-system terminology is borrowed from Camerer (2005). 16

17 2. HOMO ECONOMICUS Homo Economicus The man of economic theory, henceforth referred to as homo economicus, does not hesitate. He is rational and singleminded, and if he must pause before a decision, it is only because deciding on the optimal outcome is computationally difficult, not because he is torn apart by fundamentally different motives. He is self interested and maximizes his preferences according to well-defined mathematical rules. He is forward-looking and his preferences are consistent over long spans of time-he never finds himself gravely regretting the decision he made the day before. Finally, he clearly understands that money is just a unit of transaction, and not a commodity with inherent worth Utility maximization and Self-regarding Preferences. The homo economicus is self-interested. He doesn t care about others, but only about himself, and he walks through life maximizing his benefits and minimizing his costs. As Adam Smith writes in the Wealth of Nations, It is not from the benevolence of the butcher, the brewer, or the baker that we expect our dinner, but from their regard to their own interest. According to Smith, if the homo economicus benefits the economy and society it is only because his actions help him at the same time as well. A prime example here is the Simpsons Montgomery Burns, who when asked by Springfield s ruined citizens to share some of his electrical power, demands a list of reasons why he should, and when asked to look inside his heart, ends up releasing his hounds. If the homo economicus is asked to contribute to charity, he will say, Well, what do I get out of this? If the homo economicus finds a thick wallet lying on a sidewalk, he will happily snatch it up without giving regard to the poor chap who lost it. If homo economicus is given a pie, he will not give you a piece unless you offer a good price. If you disdainfully throw him a piece of your pie, he will shamelessly gobble it up so long as the pie is good, disregarding his seemingly beggarly position Expected Utility Theory. In the 17th century, French mathematicians Blaise Pascal and Pierre Fermat began talking about the mathematics of gambling. Their conversations led to expected value theory, which was thought to capture the essence of rational choice: the net value of a gamble, G, where amounts x 1,, x n can be won with probabilities p 1, p n is p i x i. Later, it was noticed, however, that the pleasure, or utility, derived from a certain reward did not increase linearly with the reward. For example, while most of us would derive a good amount of pleasure from winning $5 in poker, we would probably derive much less pleasure from winning those $5 after having won one million dollars in a lottery. Thus a function U was introduced by Daniel Bernoulli in the 18th century (and later axiomatized by von Neumann and Morgenstern in 1947) that measured the subjective value derived from the (monetary) reward, with the property of decreasing marginal returns captured by assuming U 0 (concavity). Instead of evaluating the expected value of the gamble G, economists today evaluate the expected utility (EU) of the gamble: p i U(x i ). By gamble, of course, I don t just mean a literal gamble of the type encountered in a casino. Rather, economists use EU theory to evaluate any decision with probabilistic outcomes. For example, when making a decision about whether or not to take a shortcut home, the expected utility of taking a shortcut is computed by multiplying the probability of success by the amount utility gained from the amount of time saved, and adding to that the probability of failure multiplied by the disutility of getting lost Consistent Preferences Over Time. Homo economicus loves himself very much, and he loves himself very much regardless of what point in time he occupies. So long as he receives his pie, he would be indifferent between receiving this pie tomorrow or today. This assumption leads to the macroeconomic concept of consumption smoothing: given $70 to spend over a week, he would choose to spend approximately $10 each day rather than spending most of it on the first day. 3 preferences are consistent over time in the sense that the same consumption path (i.e. how much he spends on each day of the week) he chooses on day 1 of the week he would also choose on any other day of the week. Holding all else constant, his preferences on day t 1 will always be the same as those on day t Money as a Unit of Exchange. This is the standard assumption in economics that the cost of a purchase is simply a reduction in the ability to make other purchases in the future. It is the basic assumption that people don t see a warm, happy glow around each dollar bill but only think of it in terms of how much it can buy. If homo economicus receives an unexpected $50 in his mailbox, he won t jump for joy until he determines what it is he that he can treat himself to that he couldn t before, and the height of his jump will be proportional to the utility he expects from that treat. When homo His 3 He wouldn t want to spend a large sum on any given day because of decreasing marginal returns, discussed in the previous section.

18 18 FROM PLATO TO NEUROECONOMICS economicus makes a purchase, he feels like he is presenting a slip of paper that documents how much is owed to him; he does not feel like he is parting with something that is real and dear to him. 3. Homo Sapiens The homo sapiens wavers before many decisions, and when he doesn t, he sometimes impulsively responds in the stupidest ways that he later regrets. When he does waver, it is usually not because he is making a decision tree and computing the expected utility, but because he is torn by anxiety, overtaken by a sudden rush of emotion, stopped still by an irrational fear. The homo sapiens is impatient, myopic about the future, and forms attachments to something as silly as green slips of paper Utility Maximization and Self-regarding Preferences. Most of us care at least a little bit about the rest of us. Furthermore, most of us prefer that everyone else cares at least a little bit about the rest of us, and we feel repulsed by characters like Montgomery Burns, who are wholly selfish. I will focus here on the ultimatum game, which is often used to study vengeance and other-regarding preferences. The ultimatum game works as follows: Two subjects are told that they will have to split $10 (or, more generally, $X) with one subject designated to be the proposer and the other the responder. The proposer proposes how the $10 will be split. For example, he could propose to keep $6 for himself and give $4 to the responder. Or he could propose to keep $9.99 for himself and give $.01 to the responder. The general proposition is of the form ($X; $10-$X), where the proser gets $X and the responder gets $10-$X. After the proposal is made, the responder responds-either he accepts the proposal, in which case he gets $10-$X and the proposer gets X, or he rejects the proposition, in which case both players get nothing. Let s consider how this game would be played by two members of the species homo economicus. The homo economicus maximizes his profits while minimizing his costs, so as the responder, he should accept any offer $10-$X, so long as the amount is positive. Just as the homo economicus would be perfectly happy to pick up any amount of money lying on the street, he would be perfectly happy to accept any amount of money offered to him by the proposer, since all he has to do is accept, at no other cost. The proposer, then, being rational and understanding of the profit-maximizing motives of his opponent and being quite selfish and profit-maximizing himself, would choose a proposal so as to maximize the amount of money with which he leaves this experiment. The proposition he would make is ($9.99; $.01), knowing full well that the responder would rather take the $.01 than get nothing. This is not what happens, however, when the experimental subjects are homo sapiens. The modal offer to the responder is $5 (Sanfey et al (2003)), and even when the responder has no power to reject the offer (also known as the Dictator Game) the proposer still finds in himself the altruistic desire to offer the responder a positive amount (Bolton [1998]). What is even more interesting is that low, but positive offers are often rejected. While a proposal of ($5; $5) has a 100% of being accepted, a proposal of ($8;$2) has an almost 50% chance of being rejected and a proposal of (9;1) has an over 60% chance of being rejected (Sanfey et al (2003)). So experimental results show the homo sapiens deviating from the homo economicus in two important ways: on the one hand, he deviates from profit maximization because of altruistic motives, and on the other, he deviates from profit maximization because of angry feelings stemming from what perceives to be a seemingly unfair and selfish offer Expected Utility Theory. Not long after von Neumann and Morgenstern formalized EU theory, experimental results began showing exceptions to the theory. Perhaps the most famous one is the Allais Paradox. In Question 1, subjects are presented with the choice between the following two gambles: Gamble 1: 100% chance of winning $1 million Gamble 2: 89% of $1 million, 10% of $10 million, 1% of nothing In Question 2, subjects are presented with the choice between the following two gambles: Gamble 1 : 89% of nothing, 11% of $1 million Gamble 2 : 90% of nothing, 10% of $5 million In Question 1, a majority of the subjects choose Gamble 1 over Gamble 2, while in Question 2, a majority of the subjects choose Gamble 2 over Gamble 1 (Kahneman (1979)). If U is the utility function with U(0) = 0, a simple mathematical calculation shows that the results of Question 1 imply that.11u(1million) >.1U(10million) whereas the results of Question 2 imply that.11u(1million) <.1U(10million). These behavioral results are obviously inconsistent with EU theory. Contradictions such as these led psychologists Kahneman and Tversky to develop a new framework for modeling decision making under uncertainty called prospect theory (Kahneman and Tversky (1979)), and a later, more complex version, called cumulative

19 4. LOOKING AT THE BRAIN 19 prospect theory (Kahneman and Tversky (1992)). The first pillar of prospect theory is that people interpret probabilities subjectively; in particular, they interpret small probabilities (larger than 0) as larger than they really, and large probabilities (smaller than 1) as smaller than they really are 4. The second pillar of prospect theory is that people are loss averse; i.e. they weight the loss of X as much heavier than the gain of X. For example, most people would reject the opportunity to play the following gamble: win $15 with probability.5, lose $10 with probability.5, even though the expected value of this gamble is $2.5 and a utility function that accounted for this phenomenon through diminishing marginal returns would have to be unrealistically concave. Experimental results have shown the coefficient of loss aversion to be 2.25 (Kahneman and Tversky (2001)); i.e., if a person stood to lose $10 with probability.5 and gain $X with probability.5, X would have to be about This type of behavior is certainly less than rational: due to an exaggerated fear of loss or failure, people reject opportunities that on average yield large benefits Consistent Preferences Over Time. Unlike the homo economicus, the homo sapiens is highly impatient and impulsive. He procrastinates, chooses cake over exercise, engages in addictive habits such as smoking, and racks up credit card fines. The general pattern of these behaviors is that the homo sapiens prefers to have a reward immediately even if that means facing a much larger cost in the future. The homo economicus would never engage in such behavior since that would imply that he cares more about his immediate than his future self. Temporal preferences are also inconsistent for the homo sapiens: Having a paper due Friday, he may put off his paper writing until Wednesday, planning on dividing the labor equally between Wednesday and Thursday. However, when Wednesday actually comes about, his preference for immediate rewards may cause him to put off all the paper writing until Thursday. Thus the Tuesday and Wednesday preferences are inconsistent with each other. We can also quantify this pattern with monetary rewards: while the homo sapiens will prefer $10 today over $11 tomorrow, he will prefer $11 in a year and one day over $10 in a year. This anomaly became known as the common difference effect (Lowenstein [1992]) in economics: a person is much more sensitive to a time delay when it occurs early rather than late. In particular, the homo sapiens is much more impatient of the time delay when the choice is between having a reward immediately and having a reward later rather than when the choice is between having a reward later and even later. This phenomenon is studied in detail by Ainslie (1975, 1992) for both human and non-human animals, and, as shown by Laibson (1997, 2001), explains consumption and investment puzzles in macroeconomics Money as a Unit of Exchange. When we look at the price of an item, do we see the price as a cost or as a painful loss? When we hoard our money, do we do it because we want to increase our future purchasing power, or because we like the feeling of sitting on our treasure chests? The answer to both questions is the latter option, because in reality people do attach emotional value to each dollar bill and do feel a pang every time they part with one. Lowenstein and Prelec (1998) give the following example of a web advertisement that plays on the attachments people make with money: Hotels are twice as good when you pay half as much. Of course, changing the price of a hotel shouldn t change the hotel itself, so the quality of a hotel should be independent of price fluctuations. What the crafty advertisers understand quite well is that the act of paying is a painful experience in itself, and so the less a customer pays, the more pleasurable his night at the hotel will be. It is for this same reason that when people pay with plastic credit cards, thus making monetary costs less salient, people tend to spend a lot more. And it is for this same reason that the ticking meter of a taxi cab makes the ride much less pleasurable. In the next section we shall see direct proof that paying is an emotional experience. 4. Looking at the Brain In just the last few years, a new discipline emerged: neuroeconomics. Economists and neuroscientist have begun a very fruitful collaboration where techniques such as functional magnetic resonance imaging (fmri) have been used to study economically relevant behaviors. In this section, I will examine what neuroscience has to say about the behaviors from Section Utility Maximization and Self-regarding Preferences. In the previous section I discussed the experimental results of the ultimatum game as an example of non-economic behavior. The next question to ask is what is actually going on in the minds of the subjects who refuse their $2 proposal. Is it anger, disgust, a rigid principle of justice, or a conditioned automatic response? Looking into a soul is very difficult, but new advances in science do allow us a peak at the brain. Sanfey 4 Showing how this subjective probability weighting function can explain the Allais Paradox is a not too difficult exercise I leave to the reader.

20 20 FROM PLATO TO NEUROECONOMICS et al (2003) decided to look inside the brains of subjects playing the ultimatum game. In their experiment, 19 participants, each in the role of the responder, were placed in an fmri scanner and asked to play the Ultimatum Game. Each subject was partnered with 10 people and told that he would play with each person once. The subjects partners were actually confederates and their offers adhered to a predetermined pattern: half of the offers were fair ($5-$5 split) and half were unfair ($9-$1, $9-$2, or $8-$3 split). The question of study was which neural structures would be engaged by the unfair offers and how they would relate to the degree of unfairness. Three structures showed greater activation by unfair offers: the anterior cingulate cortex (ACC), the dorsolateral prefrontal cortex (DLPFC), and the bilateral anterior insula. Moreover, whether an unfair offer was accepted or rejected could be determined by the relative levels of activation between the DLPFC and the anterior insula: when the DLPFC showed greater activation, the offer was accepted, whereas when the anterior insula showed greater activation, the offer was rejected. As explained by the authors, data from neuroimaging studies shows the DLPFC linked to cognitive processes such as goal maintenance and executive control. The activation of the anterior insula, on the other hand, has been linked to pain, distress, hunger, thirst, and negative emotional states such as anger and disgust. Finally, activation of the ACC has usually been linked in neuroimaging studies to instance of cognitive conflict. The conflict that this data points to is the competition between the DLPFC and the anterior insula; the competition between the cognitive goal of maximizing profits and the emotional response of disgust and aversion toward perceived unfairness Expected Utility Theory. As noted in section 2, experimental data from Kahneman and Tversky, shows that in decisions under uncertainty, people are averse to potential losses to a degree that is inconsistent with expected utility theory and greatly decreases reward accumulation in the long run. Is this deviation from rationality an emotional fear response? Shiv [2005] studied three different groups of subjects in a decision making task involving risk and potential loss. The first group consisted of patients with damage to either the ventromedial prefrontal cortex or the insular/somatosensory cortex. Both of these areas have been identified as critical to processing emotional information. The second group consisted of individuals with substance dependence (ISD). Previous studies have suggested that these individuals may have abnormalities in the aforementioned ventromedial prefrontal or insular/somatosensory cortex. Finally, the third group consisted of normal healthy individuals for control. At the beginning of the experiment, each subject was endowed with $20 and the subjects were then asked to participate in an investment task consisting of 20 rounds. In each of the 20 rounds, the subjects could either do nothing wither money or invest $1, with a 50% of losing it, and a 50% chances of winning $2.50. The appropriate decision, of course, is to invest, since the expected payoff of investing is $.75. Overall (across rounds and participants), approximately 80% of the decisions made by the ISD and lesion patients followed the optimal strategy of investing. Only about 55% of the decisions made by the control group, however, were to invest. On average, the ISD and lesion patients earned significantly more money than did the control group. The implication of these results is that the emotional responses produced by the ventromedial prefrontal and insular/somatosensory cortexes can be counterproductive when it comes to reward maximization under uncertainty because the risk/loss aversion these structures produce can sometimes be too strong Consistent Preferences Over Time. Both the ultimatum game and loss aversion can be viewed as examples of the conflict between the cognitive and emotional neural structures. McClure (2004) show that this conflict also underlies human impatience in intertemporal decision-making. In particular, they show that distinct neural systems value immediate and delayed rewards. In their experiment, subjects were placed in an fmri scanner and their brain activity was measured as they made a choice between $R at delay d and $R at delay d, where R < R and d < d. The questions were separated into two groups: d = 0 (early reward available immediately) and d > 0 (early reward is delayed), and the question of study was which neural structures would be preferentially activated by the d = 0 condition. The structures that showed much greater activation when the early reward was available immediately were the ventral striatum, medial orbitofrontal cortex, and medial prefrontal cortex; all of these are usually identified as part of the limbic system, the part of the brain that s traditionally associated with emotion, motivation, and mood. The structures that were equally activated by the d=0 and d 1 conditions (but not by non-relevant aspects of the task such as motor responses) were the right dorsolateral prefrontal cortex, right ventrolateral prefrontal cortex, and the right lateral orbitofrontal cortex; these structures have been linked to higher cognitive operations such as numerical processing and problem solving. The level of activation in these structures was proportional to the size of the reward, but mostly independent of the delay. Finally, as in the ultimatum game experiment by Sanfey et al, the subjects choices in the d = 0 could actually be predicted by looking at the structures levels of activation: when the

21 5. CONCLUSION 21 limbic structures showed greater activation than the higher cognitive structures, the smaller, immediate reward was chosen, but when the higher cognitive structures showed more activation, the larger, delayed reward was chosen. Thus impatient behavior favoring the smaller, immediate rewards stems from the emotional responses of the limbic system, whereas profit maximizing behavior that favors the larger, delayed rewards stems from the more goal-directed higher cognitive structures such as the lateral prefrontal cortex Money as a Unit of Exchange. In section 2.4, I discussed that paying, rather than being a neutral experience where paying a price is regarded as an opportunity cost of not being able to purchase other items, is probably more similar to an emotional experience where the pain of a loss (parting with money) is compared to the joy of a gain (acquiring the desired item). Knutson (2007) find paying to be largely an emotional experience when they probe the brain. In their experiment, Knutson et al offered subjects a series of purchase decisions while obtaining neuroimaging data with an fmri scanner. Each subject underwent a series of trials in which an item was first displayed, then the price, and then the offer to purchase. As in the McClure (2004) and Sanfey (2003) studies, the subjects choices could be predicted by the relative activation of distinct neural structures. It was found that in the first stage of the trial, when the product was first presented, preference for the item correlated with activation in the nucleus accumbens, whereas in the second stage, when the price was introduced, aversion to the price correlated with activity in the insula. A series of statistical tests revealed that the level of activation in the nucleus accumbens correlated very significantly with the probability of purchasing the item, whereas the level of activation in the insula correlated very significantly with the probability of not purchasing the item. As discussed in section 3.1, activity in the insula has been related to the conscious representation of emotion, such as the anticipation of pain and feelings of fear, disgust and sadness. On the other hand, anticipation of financial gains, rewards such as food or sex, or general anticipation of positive reinforcement activates the nucleus accumbens. Thus neural data does, indeed, suggest that a decision to purchase involves a competition between negative feelings of loss, on the one hand, and positive feelings of gain on the other. More generally, this study fits snugly into the pattern that can be gleaned from the neural data presented in this section: a decision does not arise from a simple set of preferences of a unified mind, but, rather, most decisions arise from a mind constantly split by various feelings, visceral desires, and higher cognitive goals. 5. Conclusion The first (extremely non-obvious) lesson that we learn is that, unlike the homo economicus, the homo sapiens has feelings. More often than not, it is emotions that determine the outcome of a decision, and not rational considerations for maximization of one s utility. These emotions can take the shape of anger and disgust as in the ultimatum game, fear of loss and uncertainty as in the investment under risk experiments, impatience and steep discounting of delayed rewards, and anticipatory feelings of loss and reward. The general trait shared by these emotional responses is that they are automatic: they need no conscious effort to be summoned, but overtake quickly and abruptly. Yet, on the other hand, as Cohen (2005) points out, the very fact that we can point to these automatic, emotional responses as sometimes being sub optimal and less than rational, means that we are perfectly capable of conceiving, and thus (in theory) acting on what the optimal or rational response should be. So it is almost as if we are of two minds: on the one hand, we feel the powerful pull of emotion but, on the other hand, something in the back of our mind asks where these emotions are really pulling us and whether that s the right direction. This is not far from the type of model neuroscience would give. Very roughly (Cohen (2005)), we can divide the brain into two parts: the neocortex, which is (relatively) evolutionarily new and is located on the outer surface of the brain, and evolutionarily older structures such as the striatum and the brainstem. Many of these older structures are either responsible for, or heavily influenced by dopamine release, which codes reward and pleasure. Thus, structures such as the ones in the brainstem respond directly and automatically to perceptions of loss and reward. In addition, these ancient structures project onto cortical regions such as the medial and orbitofrontal cortex, the amygdala, and the insula. These projections, along with their subocortical counterparts are often referred to as the limbic system, which, as previously mentioned, gives rise to emotion. Most of the evolutionary new prefrontal cortex, on the other hand, gives rise to higher cognitive functions such as problem solving abilities, mathematical reasoning, planning abilities, goal assessment, and urge suppression. Unlike the emotional responses arising from the limbic system, responses arising from the prefrontal cortex are never automatic but always effortful, deliberate, and flexible in the sense that they are highly sensitive to the particulars of a given situation. It is not surprising, then, that the limbic system and the prefrontal cortex can respond in conflicting ways

22 22 FROM PLATO TO NEUROECONOMICS to a decision. In fact, the opposing activations of the dorsolateral prefrontal cortex and the insula in the Sanfey (2003) study and the opposing activations of the orbitfrontal cortex and structures within the prefrontal cortex in the McClure (2004) study are perfect examples of the limbic-cortical conflict. This gets us back to the model of the mind that was proposed in the introduction. The human mind can viewed as the composition of two different systems: 1. the affective system and 2. the deliberative system. On the one hand, we are moved by our passions, while on the other, we are guided by the steady hand of reason. In this paper I have provided examples where these systems diverge so as to draw a clear distinction between the two. But it need not always be the case that affective responses are fundamentally irrational. In Blink: The Power of Thinking Without Thinking, Malcolm Gladwell provides many instances in which the quick emotional responses are much more helpful than the slow, deliberative ones. For example, there is the instance of the sixth century B.C. statue that after a long series of careful tests by the Getty Museum was found to be authentic, but for some unknown reason simply didn t feel right to an art historian. After much further analysis, the art historian turned out to be right. Or, there is the example of tennis coach Vic Braden who would, once in a while, would be seized by a horrible feeling as a tennis pro would serve up a ball. Every time Braden was seized by this feeling, he successfully predicted a double fault. Finally, it is also not fair to say that the deliberative system is fundamentally passionless. What Plato quite clearly recognized is that one can t give oneself to an intellectual pursuit without the right motivation, or passion. As he has Socrates relate in the Symposium, it is in pursuit of beauty that the true philosopher pursues knowledge. And, after all, wouldn t Plato s charioteer be even more aimless than his steeds if some love deep down in his heart did not endow him with a goal and a direction? We can only conclude that passion is a double-edged sword. On the one hand, it gives us motivation and cuts through difficult problems in a way that slow and effortful deliberation often can t. But on the other, it sometimes overtakes us with anger, impatience and anxiety in the most inopportune moments. Tom Petty and the Heartbreakers sing, She s gonna listen to her heart / It s gonna tell her what to do. How, and when to listen to our hearts remains an open question. References [1] Ainslie, George (1975). Specious Reward: A behavioral Theory of Impulsiveness and Impulse Control, Psychological Bulletin, LXXXII, [2] Ainslie, George (1992). Picoeconomics. Cambridge, UK Cambridge University Press, [3] Bolton, Gary E., Elena Katok, Rami Zwick (1998). International Journal of Game Theory, 27: [4] Camerer, Colin, George Loewenstein, Drazen Prelec (2003). Journal of Economic Literature, XLIII, [5] Knutson, B., Rick, S., Wimmer, G.E., Prelec, D. and Loewenstein, G. (2007). Neural Predictors of Purchases. Neuron, 53(1), [6] Laibson, David (1997). Golden Eggs and Hyperbolic Discounting. Quarterly Journal of Economic,112, [7] Loewenstein, George and Drazen Prelec (1992). Anomalies in Intertemporal Choice: Evidence and an Interpretation. The Quarterly Journal of Economics, 107 (2), [8] Lowensein, George and Drazen Prelec (1998). The Red and the Black: Mental Accounting of Savings and Debt. Marketing Science, 17(1), [9] McClure, Samuel, David Laibson, George Loewenstein and Jonathan D. Cohen, Separate Neural Systems Value Immediate and Delayed Monetary Rewards. (2004). Science, 306, [10] McClure, Samuel, Keith Ericson, David Laibson, George Loewenstein, and Jonathan Cohen (2007) Time Discounting for Primary Rewards. Journal of Neuroscience, 27: [11] Sanfey, Alan G., James K. Rilling, Jessica A. Aronson, Leigh E. Nystrom, Jonathan D. Cohen (2003). The Neural Basis of Economic Decision-Making in the Ultimatum Game Science, 300, [12] Shiv, Baba, George Loewenstein, Antoine Bechara (2005). The dark side of emotion in decision-making: When individuals with decreased emotional reactions make more advantageous decisions. Cognitive Brain Research, 23, [13] Kahneman, Daniel, and Amos Tversky (1979). Prospect Theory: An Analysis of Decision under Risk, Econometrica, XLVII (1979), Kahneman, Daniel, and Amos Tversky (1991). Loss Aversion in Riskless Choice: A Reference Dependent Model. Quarterly Journal of Economics 106, [14] Kahneman, Daniel and Amos Tversky (1992). Advances in prospect theory: Cumulative representation of uncertainty. Journal of Risk and Uncertainty, 5:297323, 1992.

23 Perelman, Poincaré, and the Ricci Flow Scott Duke Kominers 1 In July 2003, Russian mathematician Grigori Perelman announced the third and final paper of a series which solved the Poincaré Conjecture, a fundamental mathematical problem in three-dimensional topology. Perelman was awarded the Fields Medal mathematics s equivalent of the Nobel Prize for his discoveries [ICM], which were also declared the breakthrough of the year by Science [Ma]. The Poincaré Conjecture is an easily-stated, elegant question. Nonetheless, the problem stood open for nearly a century, leading mathematicians to develop more and more advanced techniques of attack. Perelman s papers [P1, P2, P3] build on the recently developed theory of the Ricci flow, extending methods pioneered by Hamilton [H1, H2, H3]. Perelman s arguments are highly innovative and technical, so much so that the mathematical community spent years evaluating and fleshing out the work. In this paper, we will introduce and set in context the topological ideas central to the Poincaré Conjecture. We focus on surfaces and their generalization, manifolds, providing the necessary intuition at each step. We will then discuss the rich history of the classification of manifolds, a topic which has driven the evolution of topology as a field of mathematics. Finally, we will summarize Perelman s work on the Ricci flow. 1. Some Relevant Topology 1.1. Surfaces. Topology begins with the familiar theory of surfaces, two-dimensional objects viewed from the perspective of three-dimensional Euclidean space. Such objects appear in mathematics as early as single-variable calculus when considering surfaces of revolution 2 and are also relevant to physics. While planes and paraboloids are surfaces, we confine our study to the properties of closed surfaces, those surfaces which are bounded in all directions. For technical reasons, we also focus our discussion on surfaces for which the concept of a clockwise rotation makes sense, the so-called orientable surfaces. 3 In closed-surface topology, any two surfaces which can be transformed into each other by continuous deformation, deformation without cutting or gluing, are considered to be equivalent. This notion is quite intuitive; in essence, it says that two surfaces are the same in some fundamental way if one can be squashed or stretched into the other in space. 4 For example, the edges of a cube can be slowly rounded out to produce a sphere and this process can be reversed to transform a sphere into a cube. More canonically, 5 a coffee cup can be transformed into a donut (or torus); each has only one handle. However, as the field of topology formalizes, it is impossible to continuously deform a sphere into a donut, as it is impossible to create a hole in the sphere without cutting. Having identified closed surfaces (such as the sphere and the donut) which are inequivalent from the perspective of topology, it is natural to ask whether one may find all equivalence classes of closed surfaces. In fact, the answer to this question has been known since the nineteenth century. Every surface S has an associated integer called the genus g(s) 0 which, in an intuitive sense, counts the number of holes in S. The genus of a sphere, then, is 0 and the genus of a donut is 1. As it turns out, the tools of topology tell us that the genus is left unchanged, or invariant, under continuous deformation. Consequently, we see again that a sphere cannot be topologically equivalent to a donut, as any continuous deformation of a sphere must result in a surface of genus 0. Even more can be said: 1 Scott Duke Kominers 09 is a mathematics concentrator and ethnomusicology minor. He studied the classification of even unimodular lattices as a 2006 PRISE Fellow and currently serves as Editor-In-Chief of The Harvard College Mathematics Review. He would like to thank Eleanor Birrell, Megan Blewett, Justin Chen, Kelley Harris, Brett Harrison, Andrea Hawksley, Paul Kominers, Stella Lee, Menyoung Lee, Daniel Litt, and Dmitry Taubinsky for helpful f and suggestions on earlier drafts of this article. He is especially grateful to Zachary Abel both for his comments and for his masterful graphic artistry. (The images in this article were typeset by Zachary Abel in Asymptote.) 2 These notions are usually introduced in order to consider the surface area formed when a given function is rotated around a line. Examples can be found in any standard calculus textbook, such as [SOP]. 3 Somewhat surprisingly, there are some surfaces which are lack a well-defined notion of left and right and are thus non-orientable. For an example, consider the Möbius strip. If one slides a clockwise-oriented arrow along a Möbius strip, the arrow becomes counterclockwise-oriented by the time it returns to its starting point. 4 A more rigorous definition of this notion may be found in any standard topology textbook. Munkres [Mu] gives an excellent introduction to the theories of metric and algebraic topology. 5 Indeed, this is a very standard example in topology. 23

24 24 PERELMAN, POINCARÉ, AND THE RICCI FLOW Theorem 1 (see [Ma1, Ma2, Mu]). Any two closed, orientable surfaces S 1 and S 2 having the same genus g(s 1 ) = g(s 2 ) are topologically equivalent. We note that, for integral g 1, an example of a genus-g surface is given by a connected sum of g donuts as shown in Figure 1. 6 Then, Theorem 1 can be rephrased as the following classification theorem: Theorem 2 (see [Ma1, Ma2, Mu]). Any closed, orientable surface is topologically equivalent to either a sphere or to a connected sum of g donuts, for some integer g 1. genus 0 genus 1 genus 2. genus 4. Figure 1. Connected sums of g donuts for g = 0, 1, 2,..., 4,... are examples of genus-g surfaces Manifolds. We observe from Theorem 2 that, topologically, a surface is essentially defined by its number of holes. Since it is easy to understand why deformation without cutting or gluing cannot affect the number of holes in a surface, this result is both intuitive and satisfying. The Poincaré Conjecture (see [Po]) is a special subcase of the Thurston Geometrization Conjecture [Th], a higherdimensional analogue of topological classification. Before we can understand higher-dimensional classification, however, we must rigorize the notion of a three-dimensional surface. For our purposes, we will define Definition 3. An n-dimensional manifold M is a topological space which is everywhere locally equivalent to Euclidean n-space R n. 7 While this definition seems quite complex, it is actually a fairly direct generalization of the types of surfaces we considered in Section The requirement that an n-manifold be a topological space is mostly technical, ensuring that we can understand the key notion of continuity, which generalizes the continuity studied in calculus. 9 The local equivalence condition in Definition 3 is somewhat harder to grasp, but can be visualized by thinking of the requirement that any person standing on a point of an n-manifold must be able to mistake the surrounding area for Euclidean n-space. Perhaps the simplest example of an n-manifold is the Euclidean n-space, itself. Further, the bounded surfaces we discussed in Section 1.1 are all two-manifolds. To see this, one need only consider how a person standing on the Earth perceives the surrounding area as flat. Indeed, at any position on a sphere, the immediately surrounding local area looks like a plane, a Euclidean two-space; similar observations lead to the conclusion that surfaces of any genus are two-manifolds. Thus, as we have claimed, the notion of a manifold directly generalizes the notion of a surface. The full n-spaces, however, are clearly unbounded as manifolds. By contrast, the bounded surfaces of Section 1.1 are bounded manifolds, that is, they are bounded in all directions. Analogously to our requirement of boundedness in Section 1.1, we will focus our discussion on bounded manifolds One might think of this connected sum as just a donut with g holes. 7 Here, we refer to ordinary Euclidean n-space, the real vector space with basis {(1, 0,..., 0),..., (0,..., 0, 1)}, which should be familiar from linear algebra. 8 In their exposition on Perelman and the Ricci flow, Gadgil and Seshadri [GS] use an equivalent definition of a manifold which is phrased in terms of gradients of smooth functions. While the gradient definition may be more appropriate to a comprehensive discussion of the Ricci flow than ours is, we have chosen to use Definition 3 for the clear geometric intuition it yields about the structure and behavior of manifolds. A significantly more rigorous formulation of the concept as we have presented it may be found in [MaT]. 9 In an ordinary calculus class, it is not uncommon to introduce ɛ-δ continuity, which topological continuity generalizes. Again, most calculus textbooks such as [SOP] explain this idea. Excellent introductions to the topological formulations of continuity can be found in [Mu] and [Sim]. 10 Owing to a shortage of mathematical terminology, there is a difference between the concept of a bounded manifold and that of a manifold with boundary, the latter of which is a manifold with a well-defined edge, such as a hollow hemisphere without a base. (In this case, the bounary is the circular edge of the hemisphere.)

25 2. THE CLASSIFICATION OF THREE-MANIFOLDS 25 D 1 γ γ Figure 2. Deformation of any simple closed curve γ on a two-sphere can reduce γ to a point, while it is impossible to deform the γ D 1 to a point. The simplest bounded n-manifold is the unit n-sphere (or just n-sphere), the set of all points (x 1,..., x n+1 ) R n+1 satisfying the equation (1) n+1 x 2 i = 1. i=1 It is clear that, for n = 2, equation (1) defines a sphere of radius 1 in Euclidean three-space, the familiar genus-0 surface which we will henceforth call the two-sphere The Classification of Three-manifolds 2.1. Preliminaries on Curves. As we saw at the end of Section 1.2, the simplest example of a three-manifold is the unit three-sphere, the set of all points (w, x, y, z) R 4 satisfying the equation (2) w 2 + x 2 + y 2 + z 2 = 1, which can be interpreted as the set of all points in Euclidean four-space of distance exactly 1 from the origin. One might think of drawing curves on a surface, by taking a pen and drawing continuous lines across the surface. Like surfaces, these curves can be continuously deformed into one another. Likewise, it is possible to study the topological properties of curves on manifolds. While this notion may seem simple, it has led to an entire field of mathematics, algebraic topology. 12 A curve on a surface is called a simple closed curve if it both 1. never crosses itself and 2. has no breaks or ends. In practice, these conditions mean that a simple closed curve is one for which it is possible to trace the curve indefinitely without crossing the curve or reaching an endpoint. For example, both circles and polygons are simple closed curves, but an arc segment on a sphere is not a simple closed curve (as it has defined endpoints). To hint at the value of simple closed curves to topology, we provide one example which will be relevant to our discussion of the Poincaré Conjecture. We begin with a definition: Definition 4. We say that an n-manifold M has the circle-shrinking property if every simple closed curve on M can be continuously deformed to a point without the curve needing to leave M. Now, if one draws an arbitrary simple closed curve γ on a two-sphere, it is possible to shrink γ to a point by continuous deformation without the curve needing to leave the sphere, as illustrated in Figure 2. By contrast, for any connected sum D g of g 1 distinct donuts, there is at least one curve γ D g which passes through a central hole. The curve γ cannot be shrunk to a point without the curve needing to leave D g, as can be clearly seen in Figure 2. Since we have seen that each surface is characterized by its number of holes, we see that the circle-shrinking property of the two-sphere identifies the two-sphere among surfaces. Indeed, every surface with holes does not have the circle-shrinking property, while every surface without holes is equivalent to the two-sphere, which does. Thus, we see that a surface has the circle-shrinking property if and only if it is topologically equivalent to the two-sphere. 11 It is worth noting the potential confusion which may result from the fact that the two-sphere resides in three-dimensional Euclidean space. This terminology arises from the fact that, for any given point on an (n 1)-sphere of radius r, the first n 1 coordinates determine the remaining coordinate. Thus, for visualization purposes, we may think of the (n 1)-sphere as being an (n 1)-dimensional manifold viewed in n-dimensional space. Nonetheless, there is some disagreement amongst mathematicians on the appropriate naming conventions for a sphere inhabiting Euclidean n-space. We use the chosen convention, however, as it is common to topologists and hence appears in all the references. 12 Munkres [Mu], Hatcher [Ha], and Massey [Ma1, Ma2] give excellent introductions to the fundamental concepts of algebraic topology. In fact, the proof of Theorems 1 and 2 depend heavily on this theory.

26 26 PERELMAN, POINCARÉ, AND THE RICCI FLOW 2.2. The Poincaré Conjecture. In the early twentieth century, Henri Poincaré [Po] noted that, like the two-sphere, the three-sphere has the circle-shrinking property. As circle-shrinking characterizes the two-sphere, Poincaré hypothesized that this property in fact characterizes the three-sphere, as well. This question later became known as the Poincaré Conjecture. 13 In modern language (paraphrased from [Mi1, Mi2]), the conjecture can be stated as follows: Conjecture 5 (The Poincaré Conjecture). If a compact three-manifold M has the circle-shrinking property, then M is topologically equivalent to the three-sphere. This seemingly simple question in the topology of three-manifolds has turned out to be extraordinarily difficult. Mathematicians made little direct progress on the conjecture until the late 1950s, when new discoveries about higher-dimensional manifolds brought about an avalanche of breakthroughs [Mi1]. In the early 1960s, Smale, Stallings, and Wallace all found proofs of analogues to the Poincaré Conjecture in dimensions five and above [Mi1, Mi2]. In the early 1980s, Freedman [Fr] showed an especially difficult four-dimensional analogue, so that the only remaining open case was Poincaré s original speculation, the conjecture in dimension three Thurston s Geometrization Conjecture. Just as Theorem 1 gives a complete classification of closed, orientable surfaces, Freedman s work [Fr] gives a complete classification of closed, four-manifolds with the circle-shrinking property. In 1983, William Thurston [Th] suggested a similar classification for three-manifolds in his far-reaching Geometrization Conjecture. Specifically, Thurston postulated that any closed three-manifold may be expressed as a connected sum of eight types of prime manifolds, which cannot be further decomposed. 15 The Poincaré Conjecture for three-manifolds arises as a special case of the Geometrization Conjecture. 16 Thus, as topologists developed new methods and insights with which to attack the Geometrization Conjecture, they implicitly worked towards a proof of the Poincaré Conjecture as well. While many mathematicians (including Thurston himself) solved special cases of the Geometrization Conjecture, the special case of manifolds of constant positive curvature which implies the Poincaré Conjecture remained intractable through the end of the twentieth century. 3. The Ricci Flow and Perelman s Proof 3.1. Hamilton s Methods. Borrowing a method from general relativity, Richard Hamilton [H1, H2, H3] attacked the Geometrization Conjecture by studying the Ricci flow, the solutions to the differential equation (3) dg ij dt = 2Ric ij, where g ij is the metric tensor and Ric ij is the Ricci curvature tensor. 17 The Ricci flow equation (3) regulates how the measurement of relative distances on a manifold changes over time. Specifically, (3) requires that distances eventually decrease in areas of positive curvature (see [Mi1, GS, An]). A generalization of the heat equation from mathematical physics, the Ricci flow equation (3) behaves as a smoothing operator in forward time but is generally impossible to reconstruct in reverse time. Like heat, which flows from hot areas to colder ones, the Ricci flow generally proceeds towards uniform curvature (see [Mi2, An]). 18 In the 1980s and 1990s, Hamilton [H1, H2, H3] developed a program which sought to characterize three-manifolds by tracing the evolution of metrics under the Ricci flow. These techniques proved to be quite fruitful; Hamilton was able to show, amongst other things, that convergence of the metric occurs in finite time for a closed three-manifold with the circle-shrinking property. Hamilton hoped to show that every metric on a closed three manifold would not only converge but would converge to a metric belonging to one of Thurston s prime manifolds. He faced problems, however, as singularities places where the metric blows up and becomes infinite develop as the metric converges. 13 Poincaré did not arrive at this hypothesis directly. Indeed, his first claim in this vein was disproven by Poincaré, himself. Milnor [Mi1, Mi2] presents excellent expositions of this and other history of three-manifold theory, much of which is outside the scope of this article. 14 See Milnor [Mi1, Mi2] for excellent summaries of these higher-dimensional results and the relevant references. 15 In simplified (but still quite mathematically advanced) language, based on Milnor s expository account [Mi2], the Geometrization Conjecture asserts: Conjecture 6 (Thurston s Geometrization Conjecture). The interior of any compact three-manifold can be split into an essentially unique connected sum of prime structures of the following forms: {The three-sphere S 3, The Euclidean space R 3, The hyperbolic space H 3, The direct product of a two-sphere and a circle, S 2 S 1, The direct product H 2 S 1 of a hyperbolic plane and a circle, A left invariant Riemannian metric on the special linear group SL 2 (R), A left invariant Riemannian metric on the Poincaré-Lorentz group, A left invariant Riemannian metric on the nilpotent Heisenberg group}. 16 Milnor [Mi1, Mi2] and Anderson [An] give explanations of the connections between the Poincaré Conjecture and the Geometrization Conjecture. 17 To directly approach the Poincaré and Geometrization conjectures with the Ricci flow techniques, it is necessary to rephrase the conjectures into forms which relate topology to Riemannian Geometry. While this leap is outside the scope of our paper, an excellent discussion can be found in Gadgil and Seshadri s exposition [GS]. 18 Several examples of explicit solutions to the Ricci flow are given by Anderson [An] and Sinestrari [Sin].

27 REFERENCES Ricci Flow with Surgery. In essence, the singularities which develop under Ricci flow were the last barrier between mathematicians and a full proof of not only the Poincaré Conjecture but of the full Geometrization Conjecture. To handle this issue, Hamilton suggested a method of Ricci flow with surgery. Conceptually, surgery means stopping the flow before a singularities develop, surgically removing and replacing the singular regions, and then restarting the flow. While somewhat simple in concept, actual application of surgery requires exact knowledge about the types of singularities which may arise. Although much progress was made at the end of the twentieth century, certain important problems remained until they were resolved by Perelman s work [P1, P2, P3]. Using incredibly insightful and inventive techniques, Perelman proved results which completely characterize the relevant structure of Ricci flow singularities. Further, his work exactly relates the flow singularities with underlying topological structures, so that no information is lost in the surgery procedure Implications. While it took several years for mathematicians to flesh out and eventually accept Perelman s assertions, the consensus is now that Perelman s proofs and methods are predominantly accurate and effective. 20 The Thurston Geometrization Conjecture, one of the most important questions in modern topology, follows directly from Perelman s results on the Ricci flow. With this, the Poincaré Conjecture for three-manifolds has finally been verified. As mentioned earlier, Perelman s achievements have garnered recognition both within and outside of the mathematical community. In the summer of 2006, Perelman was awarded the Fields Medal, mathematics s most prestigious award, for his work on the Ricci flow and the Geometrization Conjecture [ICM]. 21 The same year, the proof of the Poincaré Conjecture was proclaimed the breakthrough of the year in Science [Ma]. Further, the Clay Mathematics Institute selected the Poincaré Conjecture as one of its seven Millennium Problems in These problems are important classic questions that have resisted solution over the years, benchmarks for mathematics in the new millennium. As the first widely accepted proof of a Millennium Problem, Perelman s work on the Poincaré Conjecture could become eligible for a million-dollar prize if it continues to withstand review [CMI]. 22 The acclaim awarded to Perelman speaks to the magnitude of his accomplishment. The proof of the Poincaré Conjecture crowns nearly a century of mathematical inquiry. Further, the techniques and results of Perelman s work have forever changed the study of three-dimensional topology and opened countless doors for future research. References [An] M. T. Anderson. Geometrization of 3-manifolds and the Ricci flow. Notices of the American Mathematical Society 51 (2004), [CMI] Clay Mathematics Institute. Millennium problems May [CZ] H.-D. Cao and X.-P. Zhu. A complete proof of the Poincaré and geometrization conjectures application of the Hamilton-Perelman theory of the Ricci flow. Asian Journal of Mathematics 10 (2006), (Revised version: H.-D. Cao and X.-P. Zhu. Hamilton-Perelman s proof of the Poincaré conjecture and the geometrization conjecture. arxiv:math.dg/ , 3 Dec 2006.) [De] K. Devlin. The Millennium Problems: The Seven Greatest Unsolved Mathematical Puzzles of Our Time. Basic Books, [ICM] ICM (International Congress of Mathematicians). Fields Medal: Grigory Perelman. _info\_en.pdf, 22 Aug [Fr] M. H. Freedman. The topology of four-dimensional manifolds. Journal of Differential Geometry 17 (1982), [GS] S. Gadgil and H. Seshadri. Ricci flow and Perelman s proof of the Poincaré conjecture. Current Science 91 (2006), [H1] R. S. Hamilton. Three-manifolds with positive Ricci curvature. Journal of Differential Geometry 17 (1982), [H2] R. S. Hamilton. The formation of singularities in the Ricci flow. In: Surveys in Differential Geometry Vol. II. International Press, 1995, [H3] R. S. Hamilton. Non-singular solutions of the Ricci flow on three-manifolds. Communications in Analysis and Geometry 7 (1999), [Ha] A. Hatcher. Algebraic Topology. Cambridge University Press, [KL] B. Kleiner and J. Lott. Notes on Perelman s Papers. arxiv:math.dg/ , 21 Feb [Ma] D. Mackenzie. Breakthrough of the year: The Poincaré conjecture proved. Science 314 (2006), [MaT] I. H. Madsen and J. Tornehave. From Calculus to Cohomology: De Rham Cohomology and Characteristic Classes. Cambridge University Press, [Ma1] W. S. Massey. Algebraic Topology: An Introduction. Harcourt Brace, 1967; Springer, [Ma2] W. S. Massey. A Basic Course in Algebraic Topology. Springer, [Mi1] J. Milnor. The Poincaré Conjecture. In The Millennium Prize Problems. Clay Mathematics Institute and the American Mathematical Society, [Mi2] J. Milnor Towards the Poincaré conjecture and the classification of 3-manifolds. Notices of the American Mathematical Society 50 (2003), [MoT] J. Morgan and G. Tian. Ricci Flow and the Poincaré Conjecture. arxiv:math.dg/ , 21 Mar [Mu] J. Munkres. Topology: A First Course. Prentice-Hall, [Po] H. Poincaré. Oeuvres, Tome VI. Gauthier-Villars, [P1] G. Perelman. The entropy formula for the Ricci flow and its geometric applications. arxiv:math.dg/ , 11 Nov [P2] G. Perelman. Ricci flow with surgery on three-manifolds. arxiv:math.dg/ , 10 Mar Anderson [An] presents a very detailed explanation of the three key issues which faced mathematicians before Perelman s work. Gadgil and Seshadri [GS] give a concise outline of Perelman s proof. 20 Indeed, full expansions of Perelman s arguments have only recently been completed. Many of the details omitted from Perelman s papers were elaborated by Kleiner and Lott [KL] in a very comprehensive set of notes. An article by Cao and Zhu [CZ] recently appeared in Asian Journal of Mathematics; this first published account of Perelman s work was 366 pages, over five times the length of Perelman s original papers. More recently, Morgan and Tian [MoT] announced a book containing a self-contained proof of the Poincaré Conjecture. 21 Perelman refused the Medal, as was announced at the ceremony where he would have been presented the award. 22 Devlin [De] gives an excellent presentation of the seven Millennium Problems; his text is accessible to a nonmathematical audience.

28 28 PERELMAN, POINCARÉ, AND THE RICCI FLOW [P3] G. Perelman. Finite extinction time for the solutions to the Ricci flow on certain three-manifolds. arxiv:math.dg/ , 17 Jul [Sim] G. F. Simmons. Introduction to Topology and Modern Analysis. McGraw-Hill, [Sin] C. Sinestrari. Summer school and conference on geometry and topology of 3-manifolds: Introduction to Ricci flow. International Center for Theoretical Physics, [SOP] E. W. Swokowski, M. Olinick, and D. D. Pence. Calculus of a Single Variable. Brooks Cole, [Th] W. P. Thurston. Three-dimensional manifolds, Kleinian groups and hyperbolic geometry. In: The Mathematical Heritage of Henri Poincaré, Proceedings of Symposia in Pure Mathematics 39 (1983), Part 1. (Also in: Bulletin of the American Mathematical Society (New Series) 6 (1982), )

29 So Hot Right Now: Environmentalism is All the Rage at Harvard and Beyond Firth McEachern 1 Climate is not the only thing heating up right now-environmentalism shares the heat. In the wake of Al Gore s An Inconvenient Truth, a deluge of eco-conscious movies have hit the box office this summer, including Arctic Tale, Everything s Cool, and 11th Hour. What has caused this recent proliferation of environmental interest? Why do we suddenly value Mother Earth more? Are people responding? Environmental stewardship is not a recent phenomenon. Albeit indirectly, Arabian and Hindu cities avoided the rampant water pollution of Europe in the Middle Ages thanks to strict religious codes about cleanliness. During a time of large-scale deforestation around the Mediterranean, Ancient Greeks employed passive solar energy by orienting their homes toward the Sun. Farmer s in India, China, and Peru were wary of the dangers of soil erosion and practiced preventative measures involving crop rotations, windbreakers, and irrigation. [1] But the rise of an ideological movement, a movement that collected the myriad components of our world into a single entity dubbed the environment, is relatively recent. At the turn of the 19th century, Thomas Malthus frightened the English public with his assertion that human population would one day outstrip our resources. Among the obfuscating smog of the Industrial Revolution, increasing venereal disease rallied support for improved sanitation. Rampant disregard for the environment continued into the Gilded Age of the late 1800s, however, when use of railroads, steel, grain, and lumber proliferated, and westward expansion boomed. [2] In its wake organizations like the Sierra Club, the National Coast Anti Pollution League, and the Civilian Conservation Corps were founded. Prominent figures disagreed about the responsibilities regarding natural protection-teddy Roosevelt promulgated its wise use, while John Muir, spearhead of the first national park in the Unites States, advocated complete protection. [3] Nevertheless, respect for wilderness reached national proportions. In 1962, Rachel Carson had written her book Silent Spring, which detailed the threat of DDT, once a common insecticide, to food chains and humans, increasing the risk of cancer. [4] By this time, science had leapt into the movement, adding factual legitimacy to environmental injustices. This elevated a movement largely led by spiritually-motivated individuals who loved nature, to one informed by battalions of scientists. These scientists predicted direct and indirect perils of environmental ignorance, well highlighted by the discovery of the ozone hole in the 1980s, and more recently, global warming. Science has had an ever-prominent role in environmental policy. Wetlands were once considered nuisances, obstacles to human development and best drained. But research has proven they are incredibly intricate ecosystems, necessary for bird migration, water filtration, nutrient generation, and flood control. Advances in population dynamics and ecological theory have better outlined the threats to endangered species, and what degree of habitat protection is necessary for their survival. Science is also essential in exposing misdirected environmental policies, such as the government s demands for air quality in the lower atmosphere. [5] Ozone, although protective at high altitudes, is dangerous to humans if inhaled directly. One would assume any standards for lower ozone levels are welcoming, but the standards set for ozone levels by the government are misguided because current measures of tropospheric ozone trends are unreliable. Fortunately, the Environmental Protection Agency has begun to consider newer and better data as urged by the National Research Council. The protection of US fisheries is even more complex. Despite the establishment of a 200-mile fisheries conservation zone in 1976, fish stocks are still well below 1976 levels. [5] Improved oceanographic and biological study is needed to understand this dilemma, and ensure sustainable fish catches. With increased scientific backing, environmentalism has flourished. It has slowly shed the canard of being hijacked by a circle of tree-huggers who lack objectivity, and now garners poignant legitimacy. Fact has replaced superstition, and the public is awake. The attentive ears of Hollywood have heard the winds change, and this summer they clamored to produce what the common man is willing to hear-hence the multiplicity of movies. Admixing a dose of popular media to a serum of science has been a tour de force, helping to globalize environmental concern. Given this surge of interest, it is not so surprising that universities like Harvard have institutionalized environmental awareness. During the year, a group of a dozen-odd students (dubbed Reps ) work under the auspices of Harvard s Resource Efficiency Program (REP) to act as environmental leaders on campus. The Reps communicate with other students in their Houses and act as liaisons between students and administrators about sustainability, comfort, health, and efficiency issues. Reps teach their peers why and how 1 Firth McEachern 08 is an earth and planetary sciences concentrator and 2007 PRISE Fellow. 29

30 30 SO HOT RIGHT NOW to conserve energy, water, and materials. The Reps also suggest infrastructure and policy improvements that will remove barriers to student conservation. Student Reps focus their efforts on creating positive behavioral change, through posters, art displays, skits, workshops, and other media, with emphasis on providing factual information on consumption that is vivid and specific to the Harvard community. REP also works to educate the campus on the Harvard University Sustainability Principles declared by Harvard s former President Summers in October With University funding, REP has seen tremendous success. Electricity usage in the last fiscal year was 1.6 million kwh, or 13.8% lower than the baseline usage before REP began in Increased recycling yields saved Harvard $107,000 last year, and through increased collaboration with Habitat for Humanity, raised $75,000 from salvaged merchandise. Food waste, meanwhile, is down 32%, a difference of 180,000 lbs/yr, across the undergraduate dining halls. [6] The program has had such positive impact that a number of other universities have adopted similar frameworks. Environmentalism s benefits are twofold, and REP is case-in-point. REP s (and Harvard University s) success has not only relieved ecological stresses, but has been decidedly economic. While science has boosted political and public interest in environmental protection, sheer economic practicality has compelled its implementation. In the 19th century, writers like Henry David Thoreau praised nature s harmony; in the 20th century, maverick scientists like Alice Hamilton condemned the use of leaded fuel; and now, in the 21st century, leviathan mobilizations like the United Nations Conference on the Human Environment drew 65,000 participants to assess the situation of the globe. References [1] Kovarik, Bill. Environmental History Timeline. [2] Cronon, William. Nature s Metropolis; W. W. Norton & Company, [3] Hendricks, Stephanie. The Perverson of Wise Use (Accessed 7/20/2007) [4] Reynolds, Andy. A Brief History of Environmentalism Channel 4 Television, science/nature/environment.html [5] National Research Council. Preparing for the 21st Century: the Environment, and the Future. National Academy of Science, National Academy of Engineering, Institute of Engineering [6] Kreycik, Philip. REP successes and the Bottom Line. last updated July Numbers were extrapolated from a week-long, campus-wide food waste audit last fall.

31 Improving the Access Gap: Biological Expressions of Social Inequalities Gabriella B. Tantillo 1 In 1995, more people died of TB than in any other year in history. At least thirty million people will die from tuberculosis in the next ten years if current trends continue. Millions more will watch helplessly as friends and family members waste away, racked with coughing and sweating and fever. They may wish that medical science could cure this terrible disease. The truth is, medical science can. Since 1952, the world has had effective and powerful drugs that could make every single TB patient well again World Health Organization, 1996 According to the World Health Organization, 10 million people, most of them in developing countries, will die needlessly because they do not have access to existing medicines and vaccines. [1] With the advent of modern medicine, our ability to detect and treat infectious diseases has increased substantially. However, the disparities in disease outcome, often due to the vast differentials in access to these medical interventions, have also increased accordingly. It would appear that scientific progress has widened the gap in inequality. The global distribution of mortality due to infectious diseases follows a pattern of what anthropologists have called structural violence, which denotes the systematic ways that social structures slowly kill people by preventing them from meeting their basic needs. Life spans are reduced when people are socially dominated, politically oppressed, or economically exploited. [2] Those who occupy the lower economic rungs of society suffer increased rates of morbidity and mortality to infectious diseases, when compared to the lower death rates experienced by those who are above them on the socioeconomic ladder. As medical anthropologist Paul Farmer writes: Structural violence is visited upon all those whose social status denies them access to the fruits of scientific and social progress. [3] 1. Tuberculosis: the forgotten plague? The tuberculosis (TB) situation in the Russian penitentiary system received much media attention in the last decade. In fact, a CNN article stated that the new strain of tuberculosis is turning a stay in a Russian jail into a death sentence for many people, while a BBC article referred to the disease in Russian prison as a time bomb waiting to go off. [4, 5] In 1999, more than a million people were being held in Russian jails for a year or more in remand conditions(referring to prisoners in custody awaiting further proceedings), the waiting time it took for a trial. According to a more recent 2002 study carried out in 2 remand prisons in St. Petersburg, 876 prisoners were diagnosed with TB over the course of a three-year period. Of these, 432 were diagnosed upon entry, and 444 (over 50%) developed the disease during the course of incarceration, with the proportion diagnosed during incarceration increasing over time. The majority of cases were from young men aged less than 30 years, most of them from backgrounds of lower education and poverty. [6] In 2003, of 877,000 inmates, about 1 in 10 was shown to have TB, and about 30% of them had been infected with a drug-resistant form of the disease. Several factors contributed to the dilemmas faced. One was the growth of the prison population. Although TB incidence had always been fairly high in areas of the former Soviet Union, incidence more than doubled after 1991, with 25% of new cases occurring within the prison system. Cold, crowded, and underfunded prisons provided ideal conditions for spreading a highly infectious disease like TB, which can spread easily from inmate to inmate by simple inhalation of an infectious aerosol droplet, which released by coughing, sneezing or even talking. But what also worried health experts was the increase of a new form of TB which did not respond to standard, first-line TB drugs (rifampazin, isozianid, pyrazinamide and ethambutol). This new drug-resistant strain of TB had developed mostly due to inappropriate treatment regimes which included the widespread misuse of antibiotics. The absence of standardized medical records indicating previous treatment, as well as the low number of sputum cultures (the standard diagnostic TB test which is done by analyzing phlegm for a patient s cough) with resistance testing performed, had also contributed to the development of multi-drug resistant tuberculosis (MDRTB) during prison stays. This newer form of TB is currently much more expensive to treat, and requires more aggressive and symptomatic second-line drugs, as well as increased vigilance on the part of health care professionals. Despite these complications, MDRTB is a completely treatable condition. While some of the diseases that are currently ravaging the developing world, such as 1 Gabriella B. Tantillo 09 is a Biochemical Sciences concentrator and 2007 PRISE Fellow. She also received an inaugural Harvard Initiative for Global Health Global Infectious Diseases fellowship in She conducted research on drug resistance to HIV integrase inhibitors in the Kuritzkes Laboratory in the summer of

32 32 IMPROVING THE ACCESS GAP malaria, ebola and HIV, remain without a cure, fifty years after the introduction of almost 100% effective combination therapy to treat tuberculosis, TB remains the world s leading cause of preventable deaths in the world. In the developing world, this has become an all-too common pattern, where the newest drugs become cheaply available as generics decades after they have been saving lives in high-income countries. But why is this the case? 2. Biological Expressions of Social Inequalities Social inequalities serve to amplify the generation of drug resistant strains. In affluent environments where detection and uninterrupted treatment with quality drugs are available, such as the United States, or most other high-income nations where TB has been virtually eradicated, transmission of tuberculosis is interrupted before it can infect others. In contrast, in very poor settings, such as Africa or Haiti, where few have access to anti-tuberculosis drugs at all, the development of resistance to existing medications occurs much less frequently because it does not have the opportunity to develop in the first place. In the vast part of the developing world, those who are fortunate to start treatment for tuberculosis often fail to finish it. In most settings where tuberculosis is prevalent, TB studies performed by the World Health Organization (WHO) and other international non-governmental organizations (NGO s) have indicated a high degree of noncompliance from patients enrolled in TB treatment regimes. But how to interpret patient noncompliance? Why would an individual choose not to cooperate with a treatment regimen designed to save his or her own life? Rather than reflecting obstinacy to cooperation with medical personnel, patient noncompliance often reflects how the ability to comply with treatment regimens is significantly limited by forces that are simply beyond patient s control. Take the story of Corina Valvidia, a schoolteacher s maid, who in the 1990 s developed a debilitating cough. Her familyconsisting of her husband Carlos, her son Jaime, his wife and two grandchildren, lived together in a cramped one room house in a sprawling urban slum in Lima, Peru. [7] Given her lack of time (it took her two hours to commute to work daily on public buses) the local health clinics inconvenient hours (which coincided with her workday and would thus require her to miss work) she found it difficult to visit the local clinic. What little time she managed to have, she often spent waiting in the overbooked waiting room. She thus sought care in a private clinic, but due to increased costs, her family s small savings were soon expended, and she could not complete her treatment. As her husband recalls, they could only afford to purchase two of the four drugs that were prescribed for her treatment. She thus decided to return to the public health care system. But this brought on new complications. Now in addition to medications, she was also responsible for bringing her own medical supplies if she wanted to be treated, including syringes, gloves and gauze practices unheard of in the United States. Political unrest, including a national health care workers strike spurred by massive cuts in government spending under the new Fujimori dictatorship, further hindered her ability to attain treatment. When she finally sought treatment in a hospital, physicians curtly berated her for having waited so long to seek proper treatment, labeling her noncompliant with her regimen. Yet despite her constant efforts, none of Corina s interrupted regimes had served to improve her condition. Scarred by the condescending and ineffective treatment she received within the health-care system, Corina eventually stopped going to the hospitals. Her condition worsened, and soon she was too ill to commute to work, leaving her family scrambling for a replacement source of income. Worse, her son Jaime soon began to cough as well. For the next three years, Corina and Jaime lived with active pulmonary tuberculosis. The disease caused a strain on their family relations Jaime s wife left, leaving their two young children without a mother and under the family s care, and Corina s husband began drinking to ease the pain. When Corina began to cough up blood, she was finally was dissuaded to once again seek care; she then discovered that both she and her son (whom had likely been infected by her) were resistant to all the first-line anti-tuberculosis drugs except one. For some reason, doctors then prescribed these same medications for her again. Corina and her son of course failed to respond to these agents, and even began having threatening reactions to them. Health-care workers soon advised her to give up on her futile efforts to cure her disease. After some investigation, however, Corina s husband Carlos soon found out that other drugs second line drugs for those who had become resistant to standard-first line therapy were indeed available, but the health care system could not provide them at low cost. In fact, just two of the new drugs, ciprofloxacin and ethionamide, had a total cost of eight times Carlos monthly income. Desperate, Carlos searched high and low for ways to raise the money needed to save his wife and son. But he could not always succeed. Soon his only son was dead, leaving two grandchildren under his and his wife s care. Through the efforts of a local-community organization, Corina was finally able to obtain the drugs she needed to treat her multi-drug resistant tuberculosis (MDRTB). But by then, it was too late. Adverse side-reactions and opportunistic infections were ravaging her weakened body. Her case was deemed to severe and chronic to merit a continuation of treatment. It was not long before Corina succumbed as well, leaving the Valdivia family of Lima, Peru, completely destroyed by TB. Corina s story is sadly altogether too common in countries of middle-income like Peru. Access to second-line drugs for those who are resistant to first-line drugs are available only to those with enough money to purchase them, leading to only intermittent, interrupted access to many in need, often putting their medical treatment in jeopardy. In addition, a failure to keep accurate medical records and a severe lack of follow-up often results in patients being retreated with drugs for which they are already resistant also serves to amplify the development of drug resistance. Patients are often incapable of following a regimen to its full course due to prohibitively high treatment costs. Many national programs only offer drugs

33 5. PATENTS, LICENSING AND PUBLIC HEALTH 33 free or at low cost to patients who have tuberculosis of susceptible strains, despite the fact that the drugs for multidrug resistant tuberculosis exist and are often available. The results are structural and economic barriers that result in the unnecessary loss of human life. It is no surprise then that the vast majority of tuberculosis deaths occur in settings of crippling poverty and abysmal tuberculosis services. The settings that serve to spread drug resistance strains are those most often associated with urban poverty: crowding families, malnutrition, unhygienic living conditions, and lack of fresh air and sanitation. These variables are amplified in settings such as homeless shelters, urban slums, refugee camps and unhygienic prisons facilities. The recent outbreak of MDRTB in Siberian prisons and throughout the former Soviet Union mentioned earlier can also be attributed to these very conditions. 3. Poverty: the modern plague The plight of the impoverished caused by infectious disease are biological expressions of social inequalities. Poverty has always been the chief risk factor for acquiring and dying from tuberculosis. This was as true in the 18th and 19th century as it is today. This is true even in the United States, where economic inequality has favored tuberculosis eruptions in inner city slums, areas often already ravaged by AIDS, homelessness, intravenous drug use, and racism. In the American Journal of Public Health, Wilkinson writes: It is now clear, that the scale of income differences in a society is one of the most powerful determinants of health standards in different countries, and that it influences health through its impact on social cohesion. [8] Inequality itself, perhaps, has become the modern plague. However, even if we lack today a remedy to alleviate poverty and other social ills, we do have at our disposal the cure for almost all cases of tuberculosis. If we remain committed to championing increased access to effective drugs for TB and other diseases like HIV, we must resist restricting our field of analysis to that of economic pragmatism. Paul Farmer addresses this dilemma well in Infections and Inequalities: We are told to choose, in Haiti and in much of Africa, between treating tuberculosis and treating malnutrition. We are told to choose, in Peru, between treating those with susceptible and resistant strains. We are told to choose, in Harlem, between more funding for tuberculosis and more funding for more affordable housing. Calls for more ambitious interventions are trumped by a peculiarly bounded utilitarianism: such interventions, we are told, are not cost-effective. [9] In 1882, when Robert Koch discovered the tuberculosis bacillus, TB had been regarded not as pathology, but rather as the outcome of social misery. Koch wrote that that in the future, struggle against this dreadful plague of the human race will no longer have to contend with an indefinite something, but with an actual parasite. [10] Yet over a century later, Koch s words bear a chilling relevance. Today, tuberculosis mortality remains strikingly concentrated in setting of social and economic misery. However, this forgotten plague has become invisible now that it ceases to bother the wealthy. And tuberculosis association with extreme poverty and the third world have relegated it to irrelevance in the eyes of the affluent, which control funding for everything ranging from health-care treatment to medical research. 4. HIV: opening the world s eyes to inequality Nowhere are the inequalities of risk and the inequalities of outcome more striking than in the HIV pandemic. The leading causes of illness and death in Africa, Asia, and South America regions that account for four-fifths of the world s population are HIV/AIDS, respiratory infections, malaria, and tuberculosis. As of December 2005, of the 40.3 million people living with AIDS, 25.8 of them were in Sub-Saharan Africa, and 7.4 million in Southeast Asia, compared to only 1.2 million in North America. [11] The magnitude of the AIDS crisis in the United States and other developed nations has reawakened interest in issues of global public health. In particular, the AIDS crisis has drawn attention to the fact that millions of people in the developing world do not have access to the essential medicines that are needed to treat disease or alleviate suffering. Each day, nearly eight thousand people die of AIDS in the developing world. [12] One key factor preventing access to medicines in poor countries is the high price of new drugs; these are not deemed cost-effective in the words of health experts and policy makers. An area of fierce debate has been the price of drugs for treating serious diseases such as malaria, HIV and AIDS. Such drugs are widely available in the western world and would help manage the epidemic of these diseases in the third world. However, such drugs are hugely expensive to produce and generally well protected by patents. In order to recover the development costs, the patent holding companies charge prices for the drugs which prevent their purchase by third world countries. 5. Patents, Licensing and Public Health Several international trade agreements have attempted to adress this issue. The National Emergency Provisions of the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPs) allows the grant of compulsory licences to sell patented drugs in third world countries when they are deemed necessary for reasons of public health. However, these countries often lack the technology necessary to be able to manufacture these drugs at a low cost. Furthermore, TRIPs requires that compulsory licences only be used to fulfill a local requirement, and so it has not been possible to manufacture the drugs overseas and import them to the place of need. This creates a situation of what economists call artificial scarcity a scenario where there is a scarcity of a good even though the technology and production capacities exist to create an abundance. Prohibitive drug prices are often the result of Intellectual Property (IP) protection, which takes the form of a patent. The

34 34 IMPROVING THE ACCESS GAP owners of IP have a responsibility to ensure that their patent does not become an insurmountable barrier to appropriate health care in countries of the developing world. However, the exorbitantly high price of antiretroviral therapy the class of drugs prescribed to treat HIV/AIDS which is often several times a family s yearly income, prevent many in developing countries from obtaining these drugs. The patent system is designed to stimulate research and development (R&D) by offering the incentive of a temporary monopoly in exchange for innovation that will contribute to the advancement of human health. Monopolies allow companies to charge a price higher than their marginal cost of production, and hence earn a positive profit in the short-run. The pharmaceutical industry usually justifies high prices for medicines by pointing to the high costs of drug research and development. Drug companies purport that these higher prices are necessary in order to recover the costs of R&D, clinical trials and advertising, all crucial steps that must be undertaken in order to get drugs approved for clinical use and ready to help those in need. In practice, however, it is disproportionately the people in wealthy countries who are reaping the gains from medical progress. Ninety-seven percent of patents held worldwide are in the hands of individuals and companies in industrialized countries, and eighty percent of the patents granted in developing countries belong to residents of industrialized countries. [13] However, low and middle income countries, which make up 88.3% of the world s population, and also bear the bulk of the world s disease burden, account for only 5-7% of the overall pharmaceutical market. [14] These statistics represent a typical market failure in that these necessary medicines cannot be obtained by the populations who need them the most. Furthermore, because the development of medicines is almost entirely profit-driven, investment in research and development of the health needs and diseases endemic to third world countries, such as malaria, dengue, sleeping sickness and other tropical diseases, has come to a near standstill. [15] As a result of this differential in research investment, while common ailments in industrialized countries become increasingly treatable, many of the diseases endemic to the developing world have remained difficult to treat, as is the case in malaria, sleeping sickness, and dengue fever. When market forces fail to solve the access and R&D gap, the public sector, including international agencies, universities and public research institutes, must step in where the market fails. Thankfully, there is reason to be hopeful. 6. Closing the Access Gap In February 2001, the humanitarian organization Doctors Without Borders (Mdecins sans Frontires) sought the permission of Yale University to use a generic version of stavudine (d4t), an antiretroviral drug used to treat HIV infection, in South Africa. Doctors Without Borders made this request because it had concluded that generic stavudine could be purchased at a fraction of the cost of the branded version available in South Africa. The resulting cost savings would then allow them to expand the number of patients treated for AIDS in South Africa. Although Yale owned the patent for stavudine, they had granted an exclusive license that conferred IP rights to the pharmaceutical company Bristol-Myers Squibb. The dilemma resulted in global attention to the intense discussions that then ensued between Yale University and Bristol-Myers Squibb. Yale professor Dr. William Prusoff, who, with the late Dr. Tai-Shun Lin, demonstrated the value of stavudine in treating AIDS, stated publicly, People shouldn t die for economic reasons, because they can t afford the drug. The result of these negotiations was the first patent concession on an HIV drug resulting in a 30-fold reduction in the price of the patented drug in South Africa. This action allowed a rapid-expansion of HIV-treatment programs in Sub-Saharan Africa. [16] This event demonstrated that research institutions like Yale, as well as humanitarian organizations, and activist student groups, can play a pivotal role in closing the access gap to medical innovations. In light of this power, universities and other research institutions should be aware that their patent and licensing policies can have global ramifications. When generic makers are granted the rights to manufacture and distribute drugs, neither the universities nor the pharmaceutical companies lose, and for one simple reason: there is very little profit at stake. For 2002, Africa made up a mere 1.3% and Southeast Asia, China, and the Indian subcontinent only 6.7% of the world pharmaceutical market. [17] These markets are too small to significantly influence revenues or, unfortunately, innovation. There are however, other ways to bridge the innovation gap; for example, by licensing compounds proven effective against neglected diseases to nonprofit drug development initiatives. Last September, a student group called Universities Allied for Essential Medicines (UAEM), a coalition of students and faculty at about 25 research universities across North America, convened at Harvard University in order to discuss university actions that would improve access to medicines in poor countries. In brief, UAEM has been working to compel research universities that hold rights to new medical discoveries to only license them to branded drug makers under the condition that they be usable by generic makers in developing countries. At a recent UAEM meeting at Harvard, Dr. Jim Kim, the former AIDS director for WHO and now chairman of the Department of Social Medicine at Harvard Medical School, urged students to pressure universities to insist on making their technology accessible to the poor. 7. Conclusions We cannot separate the problem of infectious diseases from the greater social issues that surround their conception. In order to eradicate the miseries imposed by infectious disease, initiatives must also be undertaken to address the problems of poverty, gender inequality, education, and sustainable development. Likewise, there are no easy answers to the problems that often go hand-in-hand with condition of urban poverty: such as drug addiction, alcoholism, homelessness, psychiatric illness, or indifference on the part of patients. [18] However, immediate solutions to the problems of drug resistance to first-line TB

35 REFERENCES 35 drugs have been available for decades. Antiretroviral therapy has proven effective in turning AIDS from a death sentence into a chronic, but manageable condition. Medications given to pregnant and nursing mothers infected with HIV can virtually eliminate the risk of passing the disease from mother to infant through the birth canal. As of this moment, there is much that can be done. The transnational nature of the HIV and tuberculosis epidemics and other diseases demand transnational responsibility now. This responsibility rests on all of us, from the governments of nations, to international organizations, pharmaceutical companies and on academic communities at universities. At the dawn of the 21st century, the world must act with solidarity in order to contain the adversities imposed by human disease. The ultimate arbiter of success of scientific research should be its overall impact on human welfare. Only then will we be able to confidently say that the scientific research done to cure disease has not been in vain. References [1] World Health Organization. Equitable Access to essential medicines: A framework for collective action Geneva, World Health Organization (2004). [2] Gilman, Robert. Structural violence: can we find genuine peace in a world with inequitable distribution of wealth among nations? The Foundations of Peace 1997: 8. In Context 30 Dec [3] Farmer, Paul. Infections and Inequalities. University of California Press. Berkeley and Los Angeles, CA. (2001) [4] Drug-resistant TB spreading through Russian prisons. CNN.com Cable News Network. 24 June HEALTH/9906/24/russia.tb/index.html [5] World: Europe: Russian TB threatens the world. BBC.com British Broadcasting Company. 17 September co.uk/2/hi/europe/ stm [6] Lobacheva T, Sazhin V, Vdovichenko E, Giesecke J,. Pulmonary tuberculosis in two remand prisons (SIZOs) in St Petersburg, Russia. Euro Surveill 2005;10(6):93-6. [7] The full story of Corina Valdivia, as well as many other personal histories, can be found in Farmer, (2001) pp [8] Wilkinson, R.G., National Mortality Rates: The Impact of Inequality? American Journal of Public Health 82 (8): (1992) [9] Cited in Farmer, (2001) pp. 202 [10] Ibid. pp. 209 [11] UNAIDS/WHO AIDS Epidemic Update, December 2005, [12] UNAIDS, Report on the HIV/AIDS EPIDEMIC, pp. 125, 129, 133 (2000), [13] Ibid [14] UAEM Frequently Asked Questions [15] Ellen t Hoen, The Responsibility of Research Universities, The Yale Journal of Health, Policy, Law, and Ethics (Sept. 2003). [16] Dave Chokshi, Improving Access to Medicines in Poor Countries: The Role of Universities, PLos Medicine (June 2006). [pdf] [17] Amy Kapczynski et al., Global Health & University Patents, Science (Sept. 2003) [18] Ibid. pp. 249

36 Robocup 2007 Rishi Gupta 1 This summer I had the opportunity to accompany the Harvard/MIT competitive robotics club to the international Robocup competition in Atlanta, Georgia. Robocup is a week-long set of competitions and expositions revolving around autonomous robot soccer, with many dozen teams participating from forty different countries. We competed in the small size league, in which the robots play half hour games on a ping-pong table sized field with a golf ball See Fig. 1. The five robots on each side are wirelessly fed data from the teams computers and overhead cameras See Fig. 2. Though furthest from real soccer, the small size league involves the most sophisticated team work; in other leagues the robots are bipedal, play with real soccer balls, or don t have overhead cameras, but the challenges in basic movement and ball-handling preclude multi-robot plays. Figure 1. Various Views of the Robot Figure 2. Small size league robots in action There were twelve teams in the small size league this year, with widely varying skills and experience. The Harvard/MIT team was the youngest at two years old, and our goals were on the order of moving in a straight line, kicking the ball in the right direction, and recognizing the robots in our overhead camera feed. The best teams, notably CMDragons from Carnegie Mellon and Plasma-Z from Chulalongkorn Univ., Thailand, had robots that could reliably pass, block, and otherwise handle 1 Rishi Gupta, MIT 09, is a Mathematics concentrator and 2007 PRISE Fellow. 36

37 ROBOCUP the ball; in one impressive play a CMDragons robot chipped the ball over an opponent s robot to a teammate, who headbutted it past the goalie into the goal. Teams such as these could generally be scored against only with significant teamwork and precision. The most exciting part of the competition, however, was the international and collaborative aspect of the event. The teams in the small size leaugue were from Brazil, China, Finland, Germany, Japan, Mexico, Thailand, and the USA. Other leagues similarly had many countries not traditionally recognized for their engineering. Sometimes it is easy to forget that not all research is done in the handful of countries that sell products to America! Despite the language barrier and the inherently competitive nature of the event, it was not at all unusual (in PRISE collaborative spirit) for one team to put all their resources into helping another team fix a broken robot. For example, an hour before one of our matches, our opponents saw us frantically fixing a bug in our program, and not only offered to move the match to a time when our robots were working but brought over people to help us fix the problem. We ourselves donated motors to another team that ran out of spares, and sold our remaining motors to them at zero profit since they were cheaper in America than in their country. The collaboration at the event, however, was in stark contrast with the lack of synchronization in research across geographical boundaries. For instance, the East Asian countries were many years ahead of America in humanoid mechanical design, and America was many years ahead of East Asia in humanoid Artifical Intelligence. This competition has been going on for ten years, and I can t imagine what it would be like if due to language barriers and distance the only interaction between the two sides was Youtube and poorly translated papers on the Internet. One positive aspect of imperfect communication though is that it leaves room for America to develop a completely different sort of mechanical design than East Asia has, and vice-versa. It was similarly nice seeing that different countries had different parts and motors suppliers than the handful of companies that supply all of North America. Though robotics naturally lends itself to collaboration through competition, I think many fields in science and engineering could benefit from the same principle. I hope to see other such competitions starting in the future.

38 Diversity, PRISE, and an Inteview with Professor Robert A. Lue Stephanie Lo 1 Executive director of undergraduate studies. Dean of the summer school. Tutor in biochemical sciences. Senior lecturer of molecular and cellular biology. Freshman advisor. The daily obligations that accompany each position could easily surpass the time available in each day, yet Professor Robert Lue tackles each commitment with a surprising amount of energy and passion See Fig. One might expect that a professor that is so well-known not to mention, seemingly overcommitted would be inaccessible to the undergraduate population at a university as great as Harvard. Yet, as a former student of both Life Sciences 1a (LS1a) and Cell Biology (MCB54), I can attest to Lue s constant accessibility to his students, and through our conversation, I discover that Lue can sympathize with the nine hundred students in his LS1a and MCB54 classes more than we would expect. Aside from our discussion of the AP curriculum and the escalating stresses placed on high school students both issues toward which Lue is extremely sympathetic and insightful, despite how recently they have emerged Lue has not forgotten the initial struggle in determining his eventual career path. I was premed until the start of my junior year that s when I realized that what I really enjoyed was the research, Lue admits, with a small smile that is undoubtedly a manifestation of Lue s true passion for education and research. There should be a seamless continuum of research and teaching, Lue tells me. I do not view the teaching of science as something totally disengaged from research. It is no wonder, then, that Lue has played a substantial role in many of the life sciences curriculum changes seen at Harvard recently. Alluding to the growing fields of medical technology and other scientific tracks that require versatility and diversity of scientists, Lue addresses the curriculum change as both essential and overdue. With a concerned tone, he summarizes the underlying issue in one question: How can science have changed so much and the teaching of it stay the same? Enter a number of new programs, many under Lue s own watchful eye. In addition to the start of Life Sciences 1a, taught in part by Lue, PRISE was born. Despite numerous other obligations, Lue has devoted a substantial amount of time and energy to PRISE as a member of its initial steering committee. PRISE opens up a richness and a breadth that is actually not that common, Lue comments, praising the opportunities for scientists of different fields to interact. In PRISE, this diversity a word and concept that Lue continues to emphasize facilitates an experience that he feels is one of the most prominent and unique advantages of PRISE. Given the diversity of speakers during the 2007 summer - which included mathematician and former dean Benedict Gross, astrophysicist Sarah Stewart, and life scientist Erin O Shea it seems unlikely that any of the PRISE fellows were familiar with the variety of scientific fields introduced through the speakers and seminars. In addition, the student research presentations at the conclusion of PRISE have embodied diversity, with a range of titles that includes Complexity of Hardness Amplification, Building a Lego Robot Swarm, and Probing Calmodulin-Proteasome Interaction in Yeast. Even given the wonderful academic opportunities to pursue such diverse interests, Lue emphasizes the residential program of PRISE as a crucial part of creating a diverse and refined community of scientists. Referring to one of the central goals of PRISE, Lue asks, can we facilitate outside-the-classroom opportunities that engage scientists and allow them to have more of a [social] life? By the very nature of certain concentrations and lab requirements, undergraduate scientists at Harvard tend to have less time year-round than their colleagues in other areas of study; even over the course of the summer, it was not unusual for PRISE fellows to spend entire weekends in the laboratory. Still, Lue s vision for PRISE to 1 Stephanie Lo 10 is a 2007 PRISE Fellow. 38

39 INTEVIEW WITH PROFESSOR ROBERT A. LUE 39 facilitate social outings and scientific events was far from lost. This summer, nearly all of the PRISE fellows were involved in several PRISE-sponsored events, which included a whale watching boat trip, several Red Sox games, and a Tanglewood trip. Perhaps more spectacularly - albeit somewhat expected of the academically driven PRISE fellows there was a variety of student-initiated, out-of-classroom science activities, including faculty chats, a robotics group, and a PRISE abstract book. As for the future? Given the success thus far of the program, Professor Lue sees a promising future for PRISE. While there was a debate about whether PRISE should experience substantial growth in the coming years, Lue supports the maintenance of PRISE at its current size. Though he dreams that every Harvard student will be supported over the summer for research, citing considerable support already from extensive Harvard College Research Program (HCRP) and Herschel Smith awards, Lue believes that PRISE in its current condition makes a statement about diversity and community. PRISE should remain a distinct entity, Lue says, again flashing his trademark, friendly smile, PRISE should remain a jewel in the crown.

40 PRISE Horoscopes Brandon Geller 1 Leo Your rat colony grows weary of your constant injections. They stand at the ready, preparing to throw over your tyranny. The rat revolution is nigh, and when it comes, you may not be spared Virgo If you want to have a chance at being a PA next year, then you d better do your best to remain on Greg Llacer s good side. That means going to lectures, participating in activities, and, above all, NO REVEALING HIS SECRET IDENTITY, as tempting as it might be. Libra Statistics show that 78% of all PRISE fellows meet their future spouses while in PRISE. Your special someone could be that cute Physics concentrator you sat next to at dinner, or the biologist you met at the speaker series. Only one thing is sure, time is running out on finding your PRISEmate. Scorpio So what if no one else finds your topic of research interesting. It s not their topic- it s yours! Just do what you love, and those around you will accept it, and eventually, understand it. Sagittarius The stars are not very clear for you, Sagittarius. I think you need a more thorough assessment. Please report to my office in William James Hall for a full Phrenology exam if you want a true prediction. Capricorn Failure is a part of science; it just happens to be a part of science that you seem to be very good at. However, rest assured that you re no more adept at failure than the next PRISE fellow, and that your results are just as good, if not better, than could ever be expected. Aquarius Talking to your bacteria is no substitute for real people, and especially if those people are your fellow PRISE Fellows. Get out of the lab and start collaborating in really meaningful ways. Pisces You alone Pisces, of all the twelve zodiac signs, know that Dudley Herschbach speaks the truth, for you were the only one to check that when you bend over and look through your legs, everything really does sparkle. Aries Do you know what would be a more useful behavioral experiment than trying to train those mice? Training the Lev. cockroaches to stay out of our rooms, or at least to flush their own dead. Taurus The last 9 weeks have played host to the nerdiest conversations you ve ever had in your life. Topics ranging from Spice Girls to Super Mario, with Molecular Biology and Math somewhere in the middle. But admit it- you re going to miss those conversations when the summer s over. 1 Brandon Geller 08 is a biochemical sciences concentrator and 2007 PRISE Fellow.. 40