Patrick Grim

This latest volume of _The Philosopher's Annual_ presents the ten best articles published in the field during 2001. No limitations are placed on the articles' sources, subject matter or mode of treatment, providing for a diverse collection of engaging, high-caliber work that stands as a valuable sample of contemporary philosophy. This year's volume includes papers by Robert Bernasconi, Hans Halvorson, Christopher Hitchcock, Ignacio Jane, Brian Leiter, Liam Murphy and Thomas Nagel, Joel Pust, Alison Simmons, Jason Stanley and Timothy Williamson, and Crispin Wright.

Influence theory is a systematic study of formal models of the communicative influence of one person or group of people on another person or group. In that sense influence theory is an overarching philosophical discipline that includes aspects of decision theory and game theory as sub-disciplines as well as established models of de facto segregation, cultural change, opinion polarization, and epistemic networks. What we offer here is a structured outline of formal results that have been scattered across a range of disciplinary contexts from mathematics, physics and computer science to economics and political science, supplemented with a number of new models, emphasizing their place within the philosophical framework of a general theory of influence. What such an outline offers, we propose, is the prospect of new and important cross-fertilizations and expansions in formal attempts to model the diverse patterns of communicative influence.

Epistemic justifications for democracy have been offered in terms of two different aspects of decision-making: voting and deliberation, or 'votes' and 'talk.' The Condorcet Jury Theorem is appealed to as a justification in terms of votes, and the Hong-Page "Diversity Trumps Ability" result is appealed to as a justification in terms of deliberation. Both of these, however, are most plausibly construed as models of direct democracy, with full and direct participation across the population. In this paper, we explore how these results hold up if we vary the model so as to reflect the more familiar democratic structure of a representative hierarchy. We first recount extant analytic work that shows that representation inevitably weakens the voting results of the Condorcet Jury Theorem, but we question the ability of the result to shine light on real representative systems. We then show that, when we move from votes to talk, as modeled in Hong-Page, representation holds its own and even has a slight edge.

In a series of formal studies and less formal applications, Hong and Page offer a ‘diversity trumps ability’ result on the basis of a computational experiment accompanied by a mathematical theorem as explanatory background (Hong & Page 2004, 2009; Page 2007, 2011). “[W]e find that a random collection of agents drawn from a large set of limited-ability agents typically outperforms a collection of the very best agents from that same set” (2004, p. 16386). The result has been extremely influential as an epistemic justification for diversity policy initiatives. Here we show that the ‘diversity trumps ability’ result is tied to the particular random landscape used in Hong and Page’s simulation. We argue against interpreting results on that random landscape in terms of ‘ability’ or ‘expertise.’ These concepts are better modeled on smother and more realistic landscapes, but keeping other parameters the same those are landscapes on which it is groups of the best performing that do better. Smoother landscapes seem to vindicate both the concept and the value of expertise. Change in other parameters, however, also vindicates diversity. With an increase in the pool of available heuristics, diverse groups again do better. Group dynamics makes a difference as well; simultaneous ‘tournament’ deliberation in a group in place of the ‘relay’ deliberation in Hong and Page’s original model further emphasizes an advantage for diversity. ‘Tournament’ 2 dynamics particularly shows the advantage of mixed groups that include both experts and non-experts. As a whole, our modeling results suggest that relative to problem characteristics and conceptual resources, the wisdom of crowds and the wisdom of the few each have a place. We regard ours as a step toward attempting to calibrate their relative virtues in different modelled contexts of intellectual exploration.

Philosophers have long tried to understand scientific change in terms of a dynamics of revision within ‘theoretical frameworks,’ ‘disciplinary matrices,’ ‘scientific paradigms’ or ‘conceptual schemes.’ No-one, however, has made clear precisely how one might model such a conceptual scheme, nor what form change dynamics within such a structure could be expected to take. In this paper we take some first steps in applying network theory to the issue, modeling conceptual schemes as simple networks and the dynamics of change as cascades on those networks. The results allow a new understanding of two traditional approaches—Popper and Kuhn—as well as introducing the intriguing prospect of viewing scientific change using the metaphor of selforganizing criticality.

This article aims to describe the last 10 years of the collaborative scientific endeavors on polarization in particular and collective problem-solving in general by our multidisciplinary research team. We describe the team’s disciplinary composition—social psychology, political science, social philosophy/epistemology, and complex systems science— highlighting the shared and unique skill sets of our group members and how each discipline contributes to studying polarization and collective problem-solving. With an eye to the literature on team dynamics, we describe team logistics and processes that we believe make our multidisciplinary team persistent and productive. We emphasize challenges and difficulties caused by disciplinary differences in terms of terminology, units/levels of analysis, methodology, and theoretical assumptions. We then explain how work disambiguating the concepts of polarization and developing an integrative theoretical and methodological framework with complex systems perspectives has helped us overcome these challenges. We summarize the major findings that our research has produced over the past decade, and describe our current research and future directions. Last, we discuss lessons we have learned, including difficulties in a “three models” project and how we addressed them, with suggestions for effective multidisciplinary team research.

What structure of scientific communication and cooperation, between what kinds of investigators, is best positioned to lead us to the truth? Against an outline of standard philosophical characteristics and a recent turn to social epistemology, this paper surveys highlights within two strands of computational philosophy of science that attempt to work toward an answer to this question. Both strands emerge from abstract rational choice theory and the analytic tradition in philosophy of science rather than postmodern sociology of science. The first strand of computational research models the effect of communicative networks within groups, with conclusions regarding the potential benefit of limited communication. The second strand models the potential benefits of cognitive diversity within groups. Examples from each strand of research are used in analyzing what makes modeling of this sort both promising and distinctly philosophical, but are also used to emphasize possibilities for failure and inherent limitations as well.

Philosophical work on free will, contemporary as well as historical, is inevitably framed by the problem of free will and determinism. One of my goals in what follows is to give a feel for the main lines of that debate in philosophy today. I will also be outlining a particular perspective on free will. Many working philosophers consider themselves Compatibilists; the perspective outlined, building on a number of arguments in the recent literature, is a contemporary form of such a view. It cannot, however, claim to be the contemporary philosophical perspective. There is no such thing. Against a background of the perennial problem of free will and determinism, through ongoing argument and debate, philosophers continue to try to work toward an understanding of precisely what it means for an action to be free.

The purpose of this paper is to open for investigation a range of phenomena familiar from dynamical systems or chaos theory which appear in a simple fuzzy logic with the introduction of self-reference. Within that logic, self-referential sentences exhibit properties of fixed point attractors, fixed point repellers, and full chaos on the [0, 1] interval. Strange attractors and fractals appear in two dimensions in the graphing of pairs of mutually referential sentences and appear in three dimensions in the graphing of mutually referential triples.

What the Sorites has to tell us is a simple truth regarding our categories. It appears to saddle us with something other than a simple truth—something worse, a contradiction or a problem or a paradox—only when we insist on viewing it through a discrete logic of categories. Discrete categories and discrete logic are for robots. We aren’t robots, and the simple truth is that we don’t handle categories in the way any discrete logic would demand. For us non-robots, what the Sorites has to offer is a straightforward truth regarding how incapable robots and their logic are of handling categories like ours.

The iterated Prisoner’s Dilemma has become the standard model for the evolution of cooperative behavior within a community of egoistic agents, frequently cited for implications in both sociology and biology. Due primarily to the work of Axelrod (1980a, 198Ob, 1984, 1985), a strategy of tit for tat (TFT) has established a reputation as being particularly robust. Nowak and Sigmund (1992) have shown, however, that in a world of stochastic error or imperfect communication, it is not TFT that finally triumphs in an ecological model based on population percentages (Axelrod and Hamilton 1981), but ‘generous tit for tat’ (GTFT), which repays cooperation with a probability of cooperation approaching 1 but forgives defection with a probability of l/3. In this paper, we consider a spatialized instantiation of the stochastic Prisoner’s Dilemma, using two-dimensional cellular automata (Wolfram, 1984, 1986; Gutowitz, 1990) to model the spatial dynamics of populations of competing strategies. The surprising result is that in the spatial model it is not GIFT but still more generous strategies that are favored. The optimal strategy within this spatial ecology appears to be a form of ‘bending over backwards’, which returns cooperation for defection with a probability of 2/3 - a rate twice as generous as GTFT.

We work with a large spatialized array of individuals in an environment of drifting food sources and predators. The behavior of each individual is generated by its simple neural net; individuals are capable of making one of two sounds and are capable of responding to sounds from their immediate neighbors by opening their mouths or hiding. An individual whose mouth is open in the presence of food is “fed” and gains points; an individual who fails to hide when a predator is present is “hurt” by losing points. Opening mouths, hiding, and making sounds each exact an energy cost. There is no direct evolutionary gain for acts of cooperation or “successful communication” per se. In such an environment we start with a spatialized array of neural nets with randomized weights. Using standard learning algorithms, our individuals “train up” on the behavior of successful neighbors at regular intervals. Given that simple setup, will a community of neural nets evolve a simple language for signaling the presence of food and predators? With important qualifications, the answer is “yes.” In a simple spatial environment, pursuing individualistic gains and using partial training on successful neighbors, randomized neural nets can learn to communicate.

Environmental variability has been proposed as an important mechanism in behavioral psychology, in ecology and evolution, and in cultural anthropology. Here we demonstrate its importance in simulational studies as well. In earlier work we have shown the emergence of communication in a spatialized environment of wandering food sources and predators, using a variety of mechanisms for strategy change: imitation (Grim, Kokalis, Tafti & Kilb 2000), localized genetic algorithm (Grim, Kokalis, Tafti & Kilb 2001), and partial training of neural nets on the behavior of successful neighbors (Grim, St. Denis & Kokalis 2002). Here we focus on environmental variability, comparing results for all of these mechanisms in a range of different environments: (a) environments with constant resources, (b) environments with random resources around the same mean, and (c) sine-wave variable environments with cycles of ‘boom and bust’. Communication, it turns out, is strongly favored by environmental variability on the pattern of ‘boom and bust.'

The structure of communication networks can be more or less “democratic”: networks are less democratic if (a) communication is more limited in terms of characteristic degree and (b) is more tightly channeled to a few specifc nodes. Together those measures give us a two-dimensional landscape of more and less democratic networks. We track opinion volatility across that landscape: the extent to which random changes in a small percentage of binary opinions at network nodes result in wide changes across the network as a whole. If wide and frequent swings of popular opinion are taken as a mark of instability, democratic communication networks prove far more stable than anti-democratic ones. In a fnal section, we consider the democratic or anti-democratic character of networks that respond to volatility by rewiring at random, in a search for community, or in a search for a leader.

We are increasingly exposed to polarized media sources, with clear evidence that individuals choose those sources closest to their existing views. We also have a tradition of open face-to-face group discussion in town meetings, for example. There are a range of current proposals to revive the role of group meetings in democratic decision-making. Here, we build a simulation that instantiates aspects of reinforcement theory in a model of competing social influences. What can we expect in the interaction of polarized media with group interaction along the lines of town meetings? Some surprises are evident from a computational model that includes both. Deliberative group discussion can be expected to produce opinion convergence. That convergence may not, however, be a cure for extreme views polarized at opposite ends of the opinion spectrum. In a large class of cases, we show that adding the influence of group meetings in an environment of self-selected media produces not a moderate central consensus but opinion convergence at one of the extremes defined by polarized media.

The topics of modeling and information come together in at least two ways. Computational modeling and simulation play an increasingly important role in science, across disciplines from mathematics through physics to economics and political science. The philosophical questions at issue are questions as to what modeling and simulation are adding, altering, or amplifying in terms of scientific information. What changes with regard to information acquisition, theoretical development, or empirical confirmation with contemporary tools of computational modeling? In this sense the title of this article is read in the following way: What kind of information is modeling information? What kind of information does modeling give us?

Public health care interventions—regarding vaccination, obesity, and HIV, for example—standardly take the form of information dissemination across a community. But information networks can vary importantly between different ethnic communities, as can levels of trust in information from different sources. We use data from the Greater Pittsburgh Random Household Health Survey to construct models of information networks for White and Black communities--models which reflect the degree of information contact between individuals, with degrees of trust in information from various sources correlated with positions in that social network. With simple assumptions regarding belief change and social reinforcement, we use those modeled networks to build dynamic agent-based models of how information can be expected to flow and how beliefs can be expected to change across each community. With contrasting information from governmental and religious sources, the results show importantly different dynamic patterns of belief polarization within the two communities.

Beyond belief change and meme adoption, both genetics and infection have been spoken of in terms of information transfer. What we examine here, concentrating on the specific case of transfer between sub-networks, are the differences in network dynamics in these cases: the different network dynamics of germs, genes, and memes. Germs and memes, it turns out, exhibit a very different dynamics across networks. For infection, measured in terms of time to total infection, it is network type rather than degree of linkage between sub-networks that is of primary importance. For belief transfer, measured in terms of time to consensus, it is degree of linkage rather than network type that is crucial. Genes model each of these other dynamics in part, but match neither in full. For genetics, like belief transfer and unlike infection, network type makes little difference. Like infection and unlike belief, on the other hand, the dynamics of genetic information transfer within single and between linked networks are much the same. In ways both surprising and intriguing, transfer of genetic information seems to be robust across network differences crucial for the other two.

How do conventions of communication emerge? How do sounds or gestures take on a semantic meaning, and how do pragmatic conventions emerge regarding the passing of adequate, reliable, and relevant information? My colleagues and I have attempted in earlier work to extend spatialized game theory to questions of semantics. Agent-based simulations indicate that simple signaling systems emerge fairly naturally on the basis of individual information maximization in environments of wandering food sources and predators. Simple signaling emerges by means of any of various forms of updating on the behavior of immediate neighbors: imitation, localized genetic algorithms, and partial training in neural nets. Here the goal is to apply similar techniques to questions of pragmatics. The motivating idea is the same: the idea that important aspects of pragmatics, like important aspects of semantics, may fall out as a natural results of information maximization in informational networks. The attempt below is to simulate fundamental elements of the Gricean picture: in particular, to show within networks of very simple agents the emergence of behavior in accord with the Gricean maxims. What these simulations suggest is that important features of pragmatics, like important aspects of semantics, don't have to be added in a theory of informational networks. They come for free.

In order to understand the transmission of a disease across a population we will have to understand not only the dynamics of contact infection but the transfer of health-care beliefs and resulting health-care behaviors across that population. This paper is a first step in that direction, focusing on the contrasting role of linkage or isolation between sub-networks in (a) contact infection and (b) belief transfer. Using both analytical tools and agent-based simulations we show that it is the structure of a network that is primary for predicting contact infection—whether the networks or sub-networks at issue are distributed ring networks or total networks (hubs, wheels, small world, random, or scale-free for example). Measured in terms of time to total infection, degree of linkage between sub-networks plays a minor role. The case of belief is importantly different. Using a simplified model of belief reinforcement, and measuring belief transfer in terms of time to community consensus, we show that degree of linkage between sub-networks plays a major role in social communication of beliefs. Here, in contrast to the case of contract infection, network type turns out to be of relatively minor importance. What you believe travels differently. In a final section we show that the pattern of belief transfer exhibits a classic power law regardless of the type of network involved.