Grammar and Cognitive Architecture

Architecture

What is a mental grammar? We have to answer this basic question if we want to give a more detailed account of the neurocognitive structures and processes involved in language. A standard view in linguistics for the past 50 years is that a mental grammar is a task-independent and time-independent system of abstract knowledge, which should be distinguished from task-specific real-time systems that serve comprehension and production. (Sometimes the competence-performance distinction arises in this context, though I find it unhelpful.) This is a serious hypothesis, and something like this is almost certainly true for mathematical knowledge — but I believe that it is incorrect for language. I think that it’s preferable to understand the mental grammar as a real-time computational system. I also think that the evidence supports this view. This issue underlies much of our research for the past 20 years. A good orientation can be found in recent papers with Shevaun Lewis. (Phillips 1996; Phillips 2003; Phillips 2013; Lewis & Phillips 2013; Phillips & Lewis, in press)

Incremental Grammar

If the mental grammar is to be understood as a real-time computational system, does this have any impact for folks pursuing traditional linguistic questions? Sometimes, yes. In early work I argued that some fundamental problems in syntactic constituency can be solved once we assume that structures are built incrementally from left to right. The basic observation is that the word-by-word growth of a structure leads to metamorphosis: sequences that correspond to constituents at one step in the derivation cease to be constituents as new material is added. This can lead to a principled account of why different syntactic diagnostics yield apparently conflicting results (Phillips 1996, 2003). I haven’t developed this line of work further in recent years, mostly because we’ve been so consumed by other parts of the problem. There has been some subsequent sniping over specific analyses that I proposed, but to my knowledge nobody has taken on the challenge of giving a better account of constituency conflicts. A recent paper attempts a synthesis of the state of play in that domain (Phillips & Lewis 2013). A number of other linguists have presented additional arguments in favor of left-to-right grammatical derivations.

Reductionism

Some generalizations about linguistic acceptability should not be blamed on the grammar. But which ones? The literature contains a number of independent debates about whether some phenomenon should be understood as epiphenomena of language processing difficulty or semantic incoherence. The most well-known debate involves syntactic islands, whose grammatical status has been questioned almost since they were first proposed in the 1960s. But there are other less widely known debates. We should not want vague appeals to “processing difficulty” to serve as a scientific rug for sweeping recalcitrant facts under. We have studied a number of different phenomena where reductionist accounts have been proposed, and we’ve tried to develop more systematic tests and diagnostics. Each case should be taken on its merits, and our findings support or challenge reductionist accounts in different phenomena. So I find it mildly amusing that I’ve been harshly criticized for advocating a reductionist position in some cases, and harshly criticized for taking the opposite position in other cases. (Phillips 2013; Phillips 2013; Sprouse et al. 2012; Sprouse et al. 2013; Wagers et al. 2009).

Acceptability and Armchair Linguistics

One of the attractive features of linguistics is that a lot of the facts are “cheap”. You can gather dozens of acceptability judgments before breakfast and be deep into theory development by lunchtime. This means that linguists tend to spend their time worrying about different parts of the scientific process than do lab scientists. But perhaps this is all too easy, and a lot of the ‘facts’ that linguists seek to explain are bogus. This concern has been around for decades, but it has become prominent recently, in the context of the broader Repli-gate crisis in psychology and the availability of crowd-sourcing tools for internet experimentation. Some friends have argued that linguistics would be taken more seriously as a science if it replaced traditional data collection approaches with quantitative acceptability judgment methods. I disagree, and I’ve waded into the fight, sometimes with my tongue firmly in my cheek (Phillips 2010). We can’t reasonably be accused of shying away from quantitative experimental methods for studying grammar, and we test more people in large-scale acceptability studies than just about anyone (last count: 1800 participants in 50 experiments). I certainly think that linguistics has problems, but this is not the real problem, and that is not the real solution.

Theory Arbitration

Why does linguistics need the experimental methods that our team likes to use? One popular answer is that we’re useful because our new-fangled tools can provide evidence to decide among competing theoretical proposals. If traditional methods cannot settle the dispute over empty categories in unbounded dependencies, perhaps a psycholinguist can tell us the answer. And perhaps an experiment can resolve the question of whether there is structure in ellipsis sites. (Some people even refer to this as ’empirical’ evidence, contrasting with traditional ‘theoretical’ evidence. This drives me nuts. It’s all empirical!) It’s a nice idea, but I’m afraid it rarely works. Time-based methods are generally good for evaluating time-based hypotheses, and most traditional linguistic analyses disavow time-based hypotheses. In general, I think that our methods are useful not for answering questions that we already had, but for answering questions that weren’t even on our radar previously. (Phillips & Wagers 2007; Phillips & Parker 2014)

(Micro-)Variation

Every independent point of linguistic variation represents a choice point for the learner. Therefore, we should want (i) to minimize the number of choice points that learners face, and (ii) to ensure that every choice point corresponds to a readily observable fact. Otherwise, children are going to be in trouble (they’re special, but language learning is not magic). This problem formulation leads to a natural research partnership between comparative linguistics and language acquisition, and it was articulated clearly in the 1980s Principles and Parameters framework (Chomsky 1981). But the partnership was always a bit of an arms-length relationship, and more recently it has become a full-on estrangement. The 1980s program delivered little for working developmentalists, and 30 years of research on typology and micro-variation has led to despair over finding the neat parametric clusters that people dreamed of in the 1980s. I think this despair is premature, and that the field has been misled by a focus on variation in easy-to-observe properties, which dominate typological and dialect studies. It’s important to focus on understanding (i) what is “hard to observe”, and (ii) whether we find micro-variation in those hard-to-observe properties. This calls for research on children’s learning experience, and how they process and make inferences from that experience. But it also calls for a specific type of comparative linguistics, and it explains my fascination with cores-language variation in prima facie hard-to-observe phenomena such as islands, scope, and aspect. (Kush 2013; Goro 2007; Chacon et al. 2014; Kazanina & Phillips 2007)

Publications on Grammar and Cognitive Architecture

including PhD dissertations supervised

Momma, Shota (2016): Parsing, generation, and grammar . University of Maryland, 2016. (Type: PhD Thesis | Abstract | Links | BibTeX)
Chacón, Dustin Alfonso (2015): Comparative psychosyntax . University of Maryland, 2015. (Type: PhD Thesis | Abstract | Links | BibTeX)
Lewis, Shevaun; Phillips, Colin (2015): Aligning grammatical theories and language processing models . In: Journal of Psycholinguistic Research, 44 (1), pp. 27-46, 2015. (Type: Journal Article | Abstract | Links | BibTeX)
Phillips, Colin; Parker, Dan (2014): The psycholinguistics of ellipsis . In: Lingua, 151 , pp. 78-95, 2014, (published online Nov 27, 2013). (Type: Journal Article | Abstract | Links | BibTeX)
Phillips, Colin; Lewis, Shevaun (2013): Derivational order in syntax: evidence and architectural consequences . In: Studies in Linguistics, 6 , pp. 11-47, 2013. (Type: Journal Article | Abstract | Links | BibTeX)
Phillips, Colin (2013): Some arguments and non-arguments for reductionist accounts of syntactic phenomena . In: Language and Cognitive Processes, 28 , pp. 156-187, 2013. (Type: Journal Article | Abstract | Links | BibTeX)
Phillips, Colin (2013): On the nature of island constraints. II: Language learning and innateness . In: Sprouse, Jon; Hornstein, Norbert (Ed.): Experimental syntax and island effects, pp. 132-157, Cambridge University Press, 2013. (Type: Incollection | Links | BibTeX)
Phillips, Colin (2013): On the nature of island constraints. I: Language processing and reductionist accounts . In: Sprouse, Jon; Hornstein, Norbert (Ed.): Experimental syntax and island effects, pp. 64-108, Cambridge University Press, 2013. (Type: Incollection | Links | BibTeX)
Phillips, Colin (2013): Parser-grammar relations: We don’t understand everything twice . In: Sanz, Montserrat; Laka, Itziar; Tanenhaus, Michael (Ed.): Language down the garden path: the cognitive basis for linguistic structure, pp. 294–315, Oxford University Press, 2013. (Type: Incollection | Links | BibTeX)
Sprouse, Jon; Wagers, Matt; Phillips, Colin (2013): Deriving competing predictions from grammatical approaches and reductionist approaches to island effects . In: Sprouse, Jon; Hornstein, Norbert (Ed.): Experimental syntax and island effects, pp. 21–41, Cambridge University Press, 2013. (Type: Incollection | Links | BibTeX)
Sprouse, Jon; Wagers, Matt; Phillips, Colin (2012): Working memory capacity and island effects: A reminder of the issues and of the facts . In: Language, 88 , pp. 401-407, 2012. (Type: Journal Article | Abstract | Links | BibTeX)
Sprouse, Jon; Wagers, Matt; Phillips, Colin (2012): A test of the relation between working memory capacity and syntactic island effects . In: Language, 88 , pp. 82-123, 2012. (Type: Journal Article | Abstract | Links | BibTeX)
Alcocer, Pedro; Phillips, Colin (2012): Using relational syntactic constraints in content-addressable memory architectures for sentence parsing . 2012. (Type: Unpublished | Abstract | Links | BibTeX)
Phillips, Colin (2010): Should we impeach armchair linguists . In: Japanese/Korean Linguistics, 17 , pp. 49–64, 2010. (Type: Journal Article | Links | BibTeX)
Goro, Takuya (2007): Language specific constraints on scope interpretation in first language acquisition . University of Maryland, 2007. (Type: PhD Thesis | Abstract | Links | BibTeX)
Phillips, Colin; Wagers, Matthew (2007): Relating structure and time in linguistics and psycholinguistics . In: Oxford handbook of psycholinguistics, pp. 739–756, Oxford University Press, 2007. (Type: Incollection | Links | BibTeX)
Yoshida, Masaya (2006): Constraints and mechanisms in long-distance dependency formation . University of Maryland, 2006. (Type: PhD Thesis | Abstract | Links | BibTeX)
Kazanina, Nina (2005): The acquisition and processing of backwards anaphora . University of Maryland, 2005. (Type: PhD Thesis | Abstract | Links | BibTeX)
Phillips, Colin (2004): Linguistics and linking problems . In: Rice, Mabel; Warren, Steven (Ed.): Developmental Language Disorders: From Phenotypes to Etiologies, pp. 241-287, Lawrence Erlbaum Associates, Mahwah, NJ, 2004. (Type: Incollection | Links | BibTeX)
Phillips, Colin; Lau, Ellen (2004): Foundational issues . In: Journal of Linguistics, 40 (03), pp. 571–591, 2004. (Type: Journal Article | Links | BibTeX)
Phillips, Colin (2003): Linear order and constituency . In: Linguistic Inquiry, 34 (1), pp. 37–90, 2003. (Type: Journal Article | Abstract | Links | BibTeX)
Phillips, Colin; Lasnik, Howard (2003): Linguistics and empirical evidence: Reply to Edelman and Christiansen . In: Trends in Cognitive Sciences, 7 (2), pp. 61–62, 2003. (Type: Journal Article | Abstract | Links | BibTeX)
Phillips, Colin (2003): Syntax . In: Nadel, Lyn (Ed.): Encyclopedia of Cognitive Science (4), pp. 319-329, MacMillan Reference, 2003. (Type: Book Chapter | Links | BibTeX)
Phillips, Colin (2002): Constituency in deletion and movement . 2002, (Remarks on Lechner (2001)). (Type: Unpublished | Links | BibTeX)
Kim, Meesook (1999): A cross-linguistic perspective on the acquisition of locative verbs . University of Delaware, 1999. (Type: PhD Thesis | Abstract | Links | BibTeX)
Schneider, David (1999): Parsing and incrementality . University of Delaware, 1999. (Type: PhD Thesis | Abstract | Links | BibTeX)
Phillips, Colin (1998): Disagreement between adults and children . In: Mendikoetxea, Amaya; Uribe-Etxebarria, Myriam (Ed.): Theoretical Issues on the Morphology-Syntax Interface, pp. 359–394, ASJU, San Sebastian, 1998. (Type: Book Chapter | Links | BibTeX)
Phillips, Colin (1998): Teaching syntax with Trees . In: Glot International, 3 (7), 1998. (Type: Journal Article | Links | BibTeX)
Phillips, Colin (1997): Merge right: An approach to constituency conflicts . In: WCCFL XV: Proceedings of the 15th West Coast Conference on Formal Linguistics, pp. 381–395, CSLI Publications, Stanford, CA, 1997. (Type: Inproceedings | Links | BibTeX)
Phillips, Colin (1996): Order and structure . Massachusetts Institute of Technology, 1996. (Type: PhD Thesis | Abstract | Links | BibTeX)
Phillips, Colin (1996): Ergative subjects . In: Gerdts, Donna; Dziwirek, Katarzyna; Burgess, Clifford (Ed.): Grammatical Relations, CSLI Publications, Stanford, CA, 1996. (Type: Incollection | Links | BibTeX)
Phillips, Colin (1995): Right association in parsing and grammar . In: Schütze, Carson; Ganger, Jennifer; Broihier, Keven (Ed.): Papers on Language Acquisition and Proce , pp. 37–93, Department of Linguistics and Philosophy, Massachusetts Institute of Technology, vol. 26, 1995. (Type: Book Chapter | Links | BibTeX)
Phillips, Colin; Harley, Heidi (Ed.) (1994): The Morphology-syntax Connection: Proceedings of the January 1994 MIT Workshop . 1994. (Type: Collection | BibTeX)
Phillips, Colin (1994): Verbal case and the nature of polysynthetic inflection . In: Proceedings of CONSOLE 2, 1994. (Type: Inproceedings | Abstract | Links | BibTeX)
Phillips, Colin (1993): Conditions on agreement in Yimas . In: Bobaljik, Jonathan; Phillips, Colin (Ed.): Papers on Case and Agreement I , pp. 173–213, MIT Working Papers in Linguistics, vol. 18, 1993. (Type: Book Chapter | Links | BibTeX)
Bobaljik, Jonathan; Phillips, Colin (Ed.) (1993): Papers on case and agreement . vol. 18, MIT Working Papers in Linguistics, 1993. (Type: Book | BibTeX)
Phillips, Colin (Ed.) (1993): Papers on Case and Agreement II . vol. 19, MIT Working Papers in Linguistics, 1993. (Type: Book | BibTeX)