Research
You can access a smattering of recent papers (plus some ancient texts) in my Github repository. Alternatively, you can display them with the nbviewer. The README.md file (reproduced below) contains information about the specific files.
Recent papers on Agent-Based Models, populations and games
Robin Clark and Steven O. Kimbrough (2017). “Social structure, opportunistic punishment and the evolution of honest signalling”. PLOS ONE 12(12): e0188249. https://doi.org/10.1371/journal.pone.0188249
Honest signaling is generally taken to be a necessary pre-condition for a stable signaling system, because deceptive signaling at a high enough rate should cause receivers to ignore the signal, which in turn undermines the utility of sending signals. Deception is normally thought to occur because of benefits it has to the deceiver. This raises the question of why signaling systems should exist and persist over time, especially in cases in which the interests of the senders and receivers are not well aligned. Punishment has been seen as a way of imposing costs on deceptive signalers. We investigate the effects of opportunistic—that is, non-altruistic punishment—on the evolution of an honest signaling system. Our model is based on research done on social insects. We model a society of agents, divided into three castes differing in aggressiveness. Under severe punishment deception is indeed asymptotically eliminated. Under somewhat less severe punishment, deception persists and the rates of deception correlate with social structure. We find that social structure robustly mediates the level of deception under regimes of punishment and that this is evident except in the most stringent of punishment regimes.
———————————————————————————————————————————-
– Christopher Ahern, Mitchell Newberry, Robin Clark, and Joshua Plotkin. 2017. “Evolutionary forces in language change”. Nature volume 551, pages 223–226 (09 November 2017).
Both language and genes evolve by transmission over generations with opportunity for differential replication of forms. The understanding that gene frequencies change at random by genetic drift, even in the absence of natural selection, was a seminal advance in evolutionary biology. Stochastic drift must also occur in language as a result of randomness in how linguistic forms are copied between speakers. Here we quantify the strength of selection relative to stochastic drift in language evolution. We use time series derived from large corpora of annotated texts dating from the 12th to 21st centuries to analyse three well-known grammatical changes in English: the regularization of past-tense verbs, the introduction of the periphrastic ‘do’, and variation in verbal negation. We reject stochastic drift in favour of selection in some cases but not in others. In particular, we infer selection towards the irregular forms of some past-tense verbs, which is likely driven by changing frequencies of rhyming patterns over time. We show that stochastic drift is stronger for rare words, which may explain why rare forms are more prone to replacement than common ones. This work provides a method for testing selective theories of language change against a null model and reveals an underappreciated role for stochasticity in language evolution.
————————————————————————————————————————————-
-Christopher Ahern and Robin Clark (2017). “Conflict, Cheap Talk and Jespersen’s Cycle.” Semantics and Pragmatics. Volume 10, Article 11.
Game-theory has found broad application in modeling meaning in both the classical Gricean case of common interests between interlocutors and, more recently, in cases of conflicting interests. Here we consider how conflicting interests between speakers and hearers can be used to explain language change. We use tools from evolutionary game theory to characterize the effect of conflicting interests in the case of Jespersen’s cycle. We show how the cycle can be modeled as an inflationary process due to signaling with costless signals under conflicting interests. We fit the resulting dynamic model to time series data drawn from a historical corpus of Middle English.
– Robin Clark and Steven O. Kimbrough. 2017. “Hide and Seek, and the Emergence of Randomness in a Population of Deterministic Agents”. CSSS2017.
The game of Hide and Seek is interesting because it affords the study both of the emergence of coordination and anti-coordination. Socially these are important for many reasons, including modeling of the evolu- tion of identity markers and in-group–out-group tagging. We endow our agents with sensible, simple deterministic search methods, in the contex of an overall evolutionary dynamic. We find that using these methods, the agents can achieve both coordination and anti-coordination, depend- ing upon the payoffs involved. Further, we find that adding a modicum of randomization to individual search disrupts the population’s ability to coordinate. It is hardly surprising that apparently random behavior should arise from individual randomized behavior. We see this in these experiments, but we also see that it is hardly necessary. The deterministic search produced effective anti-coordination. In addition, the deterministic search produced effective coordination, in distinction to what randomized behavior achieved. Throughout, we emphasize that these are emergent population phenomena. We note as well that there is randomization occurring in this model, at the population level, if not at the individual level. These results, then, suggest that an evolutionary dynamic, acting on a population, is sufficient for apparently randomized individual behavior and that adding actual randomized individual behavior may not be necessary or even helpful.
——————————————————————————————————————————–
– Christopher Ahern, Mitchell Newberry, Robin Clark, and Joshua Plotkin. 2016. “Evolutionary forces in language change”. ms. University of Pennsylvania
Languages and genes are both transmitted from generation to generation, with opportunity for differential reproduction and survivorship of forms. Here we apply a rigorous inference framework, drawn from population genetics, to distinguish between two broad mechanisms of language change: drift and selection. Drift is change that results from stochasticity in transmission and it may occur in the absence of any in- trinsic difference between linguistic forms; whereas selection is truly an evolutionary force arising from intrinsic differences – for example, when one form is preferred by members of the population. Using large corpora of parsed texts spanning from 12th century to the 21st century, we analyze three examples of grammatical changes in English: the regularization of past-tense verbs, the rise of the periphrastic ‘do’, and syntactic variation in verbal negation. We show that we can reject stochastic drift in favor of a selective force driving some of these language changes, but not others. The strength of drift depends on a word’s frequency, and so drift provides an alternative explanation for why some words are more prone to change than others. Our results suggest an important role for stochasticity in language change, and they provide a null model against which selective theories of language evolution must be compared.
——————————————————————————————————————————————-
– Robin Clark. 2016. “Language Leaders and Effective Population Size”. ms University of Pennsylvania.
A brief look at the mathematics of effective population size; a proposal on how to accout for the data in the following paper.
—————————————————————————————————————————————-
– Robin Clark and Steven O. Kimbrough. 2015. “The Spontaneous Emergence of Language Variation from a Homogeneous Population”. CSSSA 2015. Computational Social Science Society of the Americas. (Winner best paper).
This reports on a relatively large Agent-Based Model of phonetic variation. The presence of language leaders has the effect of inducing variation.
——————————————————————————————————————————————
– Christopher Ahern and Robin Clark. 2016. “Conflict, Cheap Talk, and Jespersen’s Cycle”. (in submission).
ConflictCheapTalkandJespersensCylce.pdf
This paper uses a combination of game theory (in particular, Crawford and Sobel’s (1981) work on signaling games) and corpus data to work out a model of the change in the interpretation of negativr elements observed in many languages.
Game theory and meaning
– Robin Clark. 2016. “Games, Meaning, and Linguistic Signaling”. ms. University of Pennsylvania
A short paper on using games of incomplete information to analyze speech acts and focal points. This manuscript is to be expanded to include Bayesian updating!
—————————————————————————————————————————————–
– Christopher Ahern, Robin Clark, and Steven O. Kimbrough. 2014. “Extended Abstract: Coordination without Association”. ms, University of Pennsylvania.
cooperationwithoutAssociation.pdf
Cooperation is usually said to evolve from repeated plays, but this involves a reliable association between players. In this version, players might play one-shot, but are given (some) information about the average return of (a subset of) other palyers. They all strive to be above average in their earnings; if they earn below average, they adjust their play, with the result that a certain level of cooperation is always reliably present and could act as a ratchet for the evolution of cooperation.
—————————————————————————————————————————————–
– Robin Clark. 2012. “Social and Physical Coordination”. Interaction Studies. 13(1). 66-79.
socialandPhysicalCoordination.pdf
As the title suggests, a short survey on connections between social coordination (groups of people coordinating their behavior) and physical coordination (how we, for example, coordinate body movements).
——————————————————————————————————————————————
– Robin Clark and Prashant Parikh. 2007. “Game Theory and Discourse Anaphora”. Journal of Logic, Language and Information, 16, 265-282.
gameTheoryandDiscourseAnaphora.pdf
This is a model of reference tracking using game theoretic contructs. I think I would more carefully incorporate focal points and updating; some of this was done in Meaningful Games (2012; The MIT Press).
——————————————————————————————————————————————
– Laia Mayol and Robin Clark. 2010. “Pronouns in Catalan: Games of partial information and the use of linguistic resources.” Journal of Pragmatics. 42. 781-799.
Journal of Pragmatics 2010 Mayol.pdf
This paper uses the framework of Clark and Parikh (2007) to look at Catalan; interesting because Catalan has null pronouns. In other words, Catalan has a resource that English lacks and is able to distribute its resources differently!
——————————————————————————————————————————————
– Robin Clark. 2009. “Games, Quantification, and Discourse Structure” in O. Majer, A-V Pietarinen and T. Tulenheimo (ends). Games: Unifying Logic, Language, and Philosophy. Springer. p. 139-150.
Clark_2009_Games_Quantification_DS.pdf
This paper gives a simple game theoretic method of evaluating logical quantifiers and updating the discourse model.
Neuroscience: Quantifiers, Number Sense and Strategic Reasoning
– Robin Clark. 2011. “Generalized quantifiers and number sense”. Philosophy Compass. 6/9. 611-621.
generalizedQuantifiersandNumberSense.pdf
An overview of the relationship between quantifiers, semantic automata, and numerosity.
—————————————————————————————————————————————–
– Stefan Heim, Corey T. McMillan, Robin Clark, Laura Baehr, Kylie Ternes, Christopher Olm, Nam Eun Min, and Murray Grossman. 2016. “How the brain learns how few are ‘many’: An fMRI study of the flexibility of quantifier semantics”. NeuroImage. 125, 45-52.
This looks at the interpretation of “vague” quantifiers like few and many and illustrates how their interpretation can be manipulated by (slight) training.
– Nicola Spotorno, Meghan Healey, Corey T. McMillan, Katya Rascovsky, David J. Irwin, Robin Clark and Murray Grossman. 2015. “Processing Ambiguity in a Linguistic Context: Decision-Making Difficulties in Non-Aphasic Patients with Behavioral Variant Frontotemporal Degeneration.” Frontiers in Human Neuroscience 9 (October): 1–8. doi:10.3389/fnhum.2015.00583.
Front. Hum. Neurosci. 2015 Spotorno.pdf
Some extent of ambiguity is ubiquitous in everyday conversations. For example, words have multiple meaning and very common pronouns, like “he” and “she” (anaphoric pronouns), have little meaning on their own and refer to a noun that has been previously introduced in the discourse. Ambiguity triggers a decision process that is not a subroutine of language processing but rather a more general domain resource. Therefore non-aphasic patients with limited decision-making capability can encounter severe limitation in language processing due to extra linguistic limitations. In the present study, we test patients with behavioral variant frontotemporal degeneration (bvFTD), focusing on anaphora as a paradigmatic example of ambiguity resolution in the linguistic domain.
——————————————————————————————————————————————–
– Robin Clark and Murray Grossman. 2007. “Number Sense and Quantifier Interpretation.” Topoi 26 (1): 51–62. doi:10.1007/s11245-006-9008-2.
We consider connections between number sense—the ability to judge number—and the inter- pretation of natural language quantifiers. In particular, we present empirical evidence concerning the neuro- anatomical underpinnings of number sense and quan- tifier interpretation. We show, further, that impairment of number sense in patients can result in the impair- ment of the ability to interpret sentences containing quantifiers. This result demonstrates that number sense supports some aspects of the language faculty.
——————————————————————————————————————————————–
– Corey T McMillan, Robin Clark, Peachie Moore, and Murray Grossman. 2006. “Quantifier Comprehension in Corticobasal Degeneration.” Brain and Cognition 62 (3): 250–60. doi:10.1016/j.bandc.2006.06.005.
Brain and Cognition 2006 Mcmillan.pdf
In this study, we investigated patients with focal neurodegenerative diseases to examine a formal linguistic distinction between classes of generalized quantifiers, like “some X” and “less than half of X.” Our model of quantifier comprehension proposes that number knowledge is required to understand both first-order and higher-order quantifiers. The present results demonstrate that corticobasal degeneration (CBD) patients, who have number knowledge impairments but little evidence for a deficit understanding other aspects of language, are impaired in their comprehension of quantiWers relative to healthy seniors, Alzheimer’s disease (AD) and frontotemporal dementia (FTD) patients [F(3, 77) D 4.98; p < .005]. Moreover, our model attempts to honor a distinction in complexity between classes of quantifiers such that working memory is required to comprehend higher-order quantifiers. Our results support this distinction by demonstrating that FTD and AD patients, who have working memory limitations, have greater difficulty understanding higher-order quantifiers relative to first-order quantiWers [F(1, 77) D 124.29; p < .001]. An important implication of these findings is that the meaning of generalized quantifiers appears to involve two dissociable components, number knowledge and working memory, which are supported by distinct brain regions.
——————————————————————————————————————————————-
– Corey T McMillan, Robin Clark, Peachie Moore, Christian Devita, and Murray Grossman. 2005. “Neural Basis for Generalized Quantifier Comprehension.” Neuropsychologia 43 (12): 1729–37. doi:10.1016/j.neuropsychologia.2005.02.012.
Neuropsychologia 2005 Mcmillan.pdf
Generalized quantifiers like “all cars” are semantically well understood, yet we know little about their neural representation. Our model of quantifier processing includes a numerosity device, operations that combine number elements and working memory. Semantic theory posits two types of quantifiers: first-order quantifiers identify a number state (e.g. “at least 3”) and higher-order quantifiers additionally require maintaining a number state actively in working memory for comparison with another state (e.g. “less than half”). We used BOLD fMRI to test the hypothesis that all quantifiers recruit inferior parietal cortex associated with numerosity, while only higher-order quantifiers recruit prefrontal cortex associated with executive resources like working memory. Our findings showed that first-order and higher-order quantifiers both recruit right inferior parietal cortex, suggesting that a numerosity component contributes to quantifier comprehension. Moreover, only probes of higher-order quantifiers recruited right dorsolateral prefrontal cortex, suggesting involvement of executive resources like working memory. We also observed activation of thalamus and anterior cingulate that may be associated with selective attention. Our findings are consistent with a large-scale neural network centered in frontal and parietal cortex that supports comprehension of generalized quantifiers.
Old papers on learnability and language change
– Robin Clark. 1992. “The Selection of Syntactic Knowledge.” Language Acquisition. 2(2). 83-149.
selectionofSyntacticKnowledge.pdf
An experiment in using Genetic Algorithms to do parameter setting. Also shows an early interest in evolution and population modeling!
———————————————————————————————————————————————
– Robin Clark and Ian Roberts. 1993. “A Computational Model of Language Learnability and Language Change.” Linguistic Inquiry. 24(2). 299-345.
computationalModelofLanguageLearnabilityandLanguageChange.pdf
This paper tries to apply the learning model of “The Selection of Syntactic Knowledge” to a case of language change (the position of verbs in French).