Across sign languages, nouns can be derived from verbs through morphophonological changes in movement by (1) movement reduplication and size reduction or (2) size reduction alone. We asked whether these cross-linguistic similarities arise from cognitive biases in how humans construe objects and actions. We tested nonsigners' sensitivity to differences in noun-verb pairs in American Sign Language (ASL) by asking MTurk workers to match images of actions and objects to videos of ASL noun-verb pairs. Experiment 1a's match-to-sample paradigm revealed that nonsigners interpreted all signs, regardless of lexical class, as actions. The remaining experiments used a forced-matching procedure to avoid this bias. Counter our predictions, nonsigners associated reduplicated movement with actions not objects (inversing the sign language pattern) and exhibited a minimal bias to associate large movements with actions (as found in sign languages). Whether signs had pantomimic iconicity did not alter nonsigners' judgments. We speculate that the morphophonological distinctions in noun-verb pairs observed in sign languages did not emerge as a result of cognitive biases, but rather as a result of the linguistic pressures of a growing lexicon and the use of space for verbal morphology. Such pressures may override an initial bias to map reduplicated movement to actions, but nevertheless reflect new iconic mappings shaped by linguistic and cognitive experiences.

Iconicity is often defined as the resemblance between a form and a given meaning, while transparency is defined as the ability to infer a given meaning based on the form. This study examined the influence of knowledge of American Sign Language (ASL) on the perceived iconicity of signs and the relationship between iconicity, transparency (correctly guessed signs), 'perceived transparency' (transparency ratings of the guesses), and 'semantic potential' (the diversity (H index) of guesses). Experiment 1 compared iconicity ratings by deaf ASL signers and hearing non-signers for 991 signs from the ASL-LEX database. Signers and non-signers' ratings were highly correlated; however, the groups provided different iconicity ratings for subclasses of signs: nouns vs. verbs, handling vs. entity, and one- vs. two-handed signs. In Experiment 2, non-signers guessed the meaning of 430 signs and rated them for how transparent their guessed meaning would be for others. Only 10% of guesses were correct. Iconicity ratings correlated with transparency (correct guesses), perceived transparency ratings, and semantic potential (H index). Further, some iconic signs were perceived as non-transparent and vice versa. The study demonstrates that linguistic knowledge mediates perceived iconicity distinctly from gesture and highlights critical distinctions between iconicity, transparency (perceived and objective), and semantic potential.

Our research on statistical language learning shows that infants, young children, and adults can compute, online and with remarkable speed, how consistently sounds co-occur, how frequently words occur in similar contexts, and the like, and can utilize these statistics to find candidate words in a speech stream, discover grammatical categories, and acquire simple syntactic structure in miniature languages. However, statistical learning is not merely learning the patterns presented in the input. When their input is inconsistent, children sharpen these statistics and produce a more systematic language than the one to which they are exposed. When input languages inconsistently violate tendencies that are widespread in human languages, learners shift these languages to be more aligned with language universals, and children do so much more than adults. These processes explain why children acquire language (and other patterns) more effectively than adults, and also may explain how systematic language structures emerge in communities where usages are varied and inconsistent. Most especially, they suggest that usage-based learning approaches must account for differences between adults and children in how usage properties are acquired, and must also account for substantial changes made by adult and child learners in how input usage properties are represented during learning.

Embodied theories of cognition propose that humans use sensorimotor systems in processing language. The Action-Sentence Compatibility Effect (ACE) refers to the finding that motor responses are facilitated after comprehending sentences that imply movement in the same direction. In sign languages there is a potential conflict between sensorimotor systems and linguistic semantics: movement away from the signer is perceived as motion toward the comprehender. We examined whether perceptual processing of sign movement or verb semantics modulate the ACE. Deaf ASL signers performed a semantic judgment task while viewing signed sentences expressing toward or away motion. We found a significant congruency effect relative to the verb's semantics rather than to the perceived motion. This result indicates that (a) the motor system is involved in the comprehension of a visual-manual language, and (b) motor simulations for sign language are modulated by verb semantics rather than by the perceived visual motion of the hands.

Motor theories of perception posit that motor information is necessary for successful recognition of actions. Perhaps the most well known of this class of proposals is the motor theory of speech perception, which argues that speech recognition is fundamentally a process of identifying the articulatory gestures (i.e. motor representations) that were used to produce the speech signal. Here we review neuropsychological evidence from patients with damage to the motor system, in the context of motor theories of perception applied to both manual actions and speech. Motor theories of perception predict that patients with motor impairments will have impairments for action recognition. Contrary to that prediction, the available neuropsychological evidence indicates that recognition can be spared despite profound impairments to production. These data falsify strong forms of the motor theory of perception, and frame new questions about the dynamical interactions that govern how information is exchanged between input and output systems.

I suggest two puzzles for the Mirror System Hypothesis. First, there is little evidence that mirror neuron populations for words or for signs exist in Broca's area, and a mirror system is not critical for either speech or sign perception. Damage to Broca's area (or to the mirror system for human action) does not result in deficits in sign or speech perception. Second, the gesticulations of speakers are highly integrated with speech, but pantomimes and modern protosigns (conventional gestures) are not co-expressive with speech, and they do not co-occur with speech. Further, signers also produce global, imagistic gesticulations with their mouths and bodies simultaneously while signing with their hands. The expanding spiral of protosign and protospeech does not predict the integrated and co-expressive nature of modern gestures produced by signers and speakers.

Duality of Patterning, one of Hockett's (1960) proposed design features unique to human language, refers in part to the arrangements of a relatively small stock of distinguishable meaningless sounds which are combined to create a potentially infinite set of morphemes. Literature regarding the emergence of this design feature is less abundant than that exploring other levels of structure as focus is more often given to the emergence of syntax. In an effort to explore where combinatorial structure of meaningless elements arises the results of two pilot experiments are presented within which we observe human participants modifying a small lexicon of visual symbols through a process of iterated learning. As this lexicon evolves there is evidence that it becomes simpler and more learnable, more easily transmitted. I argue that these features are a consequence of spontaneous emergence of combinatorial, sub-lexical structure in the lexicon, that the pattern of emergence is more complex than the most widely espoused explanation suggests, and I propose ways in which future work can build on what we learn from these pilot experiments to confirm this hypothesis.

In human speech, a finite set of basic sounds is combined into a (potentially) unlimited set of well-formed morphemes. Hockett (1960) placed this phenomenon under the term 'duality of patterning' and included it as one of the basic design features of human language. Of the thirteen basic design features Hockett proposed, duality of patterning is the least studied and it is still unclear how it evolved in language. Recent work shedding light on this is summarized in this paper and experimental data is presented. This data shows that combinatorial structure can emerge in an artificial whistled language through cultural transmission as an adaptation to human cognitive biases and learning. In this work the method of experimental iterated learning (Kirby et al. 2008) is used, in which a participant is trained on the reproductions of the utterances the previous participant learned. Participants learn and recall a system of sounds that are produced with a slide whistle. Transmission from participant to participant causes the whistle systems to change and become more learnable and more structured. These findings follow from qualitative observations, quantitative measures and a follow-up experiment that tests how well participants can learn the emerged whistled languages by generalizing from a few examples.

Adult speakers use verbs in syntactically appropriate ways. For example, they know implicitly that the boy hit at the fence is acceptable but the boy broke at the fence is not. We suggest that this knowledge is lexically encoded in semantic decompositions. The decomposition for break verbs (e.g. crack, smash) is hypothesized to be more complex than that for hit verbs (e.g. kick, kiss). Specifically, the decomposition of a break verb denotes that "an entity changes state as the result of some external force" whereas the decomposition for a hit verb denotes only that "an entity potentially comes in contact with another entity." In this article, verbs of the two types were compared in a lexical decision experiment - Experiment 1 - and they were compared in sentence comprehension experiments with transitive sentences (e.g. the car hit the bicycle and the car broke the bicycle) - Experiments 2 and 3. In Experiment 1, processing times were shorter for the hit than the break verbs and in Experiments 2 and 3, processing times were shorter for the hit sentences than the break sentences, results that are in accord with the complexities of the postulated semantic decompositions.

Speakers move their hands when they talk--they gesture. These gestures can signal whether the speaker is ready to learn a particular task and, in this sense, provide a window onto the speaker's knowledge. But gesture can do more than reflect knowledge. It can play a role in changing knowledge in at least two ways: indirectly through its effects on communication with the learner, and directly through its effects on the learner's cognition. Gesturing is, however, not limited to learners. Speakers who are proficient in a task also gesture. Their gestures have a different relation to speech than the gestures that novices produce, and seem to support cognition rather than change it. Gesturing can thus serve as a tool for thinking and for learning.

The idea that concepts are embodied by our motor and sensory systems is popular in current theorizing about cognition. Embodied cognition accounts come in different versions and are often contrasted with a purely symbolic amodal view of cognition. Simulation, or the hypothesis that concepts simulate the sensory and motor experience of real world encounters with instances of those concepts, has been prominent in psychology and cognitive neuroscience. Here, with a focus on spatial thought and language, I review some of the evidence cited in support of simulation versions of embodied cognition accounts. While these data are extremely interesting and many of the experiments are elegant, knowing how to best interpret the results is often far from clear. I point out that a quick acceptance of embodied accounts runs the danger of ignoring alternate hypotheses and not scrutinizing neuroscience data critically. I also review recent work from my lab that raises questions about the nature of sensory motor grounding in spatial thought and language. In my view, the question of whether or not cognition is grounded is more fruitfully replaced by questions about gradations in this grounding. A focus on disembodying cognition, or on graded grounding, opens the way to think about how humans abstract. Within neuroscience, I propose that three functional anatomic axes help frame questions about the graded nature of grounded cognition. First, are questions of laterality differences. Do association cortices in both hemispheres instantiate the same kind of sensory or motor information? Second, are questions about ventral dorsal axes. Do neuronal ensembles along this axis shift from conceptual representations of objects to the relationships between objects? Third, are questions about gradients centripetally from sensory and motor cortices towards and within perisylvian cortices. How does sensory and perceptual information become more language-like and then get transformed into language proper?

The ontological distinction between discrete individuated objects and continuous substances, and the way this distinction is expressed in different languages has been a fertile area for examining the relation between language and thought. In this paper we combine simulations and a cross-linguistic word learning task as a way to gain insight into the nature of the learning mechanisms involved in word learning. First, we look at the effect of the different correlational structures on novel generalizations with two kinds of learning tasks implemented in neural networks-prediction and correlation. Second, we look at English- and Spanish-speaking 2-3-year-olds' novel noun generalizations, and find that count/mass syntax has a stronger effect on Spanish- than on English-speaking children's novel noun generalizations, consistent with the predicting networks. The results suggest that it is not just the correlational structure of different linguistic cues that will determine how they are learned, but the specific learning mechanism and task in which they are involved.