Marco Mazzone

Concepts: Stored or Created?

Are concepts stable entities, unchanged from context to context? Or rather are they context-dependent structures, created on the fly? We argue that this does not constitute a genuine dilemma. Our main thesis is that the more a pattern of features is general and shared, the more it qualifies as a concept. Contextualists have not shown that conceptual structures lack a stable, general core, acting as an attractor on idiosyncratic information. What they have done instead is to give a contribution to the comprehension of how conceptual structure organized around such a stable core can produce contextually appropriate representations on demand.

First Words Learning: A Cortical Model

Humans come to recognize an infinite variety of natural and man-made objects in their lifetime and make use of sounds to identify and categorize them. How does this lifelong learning process begin? Many hypotheses have been proposed to explain the learning of first words with some emerging from the particular characteristics observed in child development. One is the peculiar trend in the speed with which words are learned, which have been referred to in the literature as “fast mapping”. We present a neural network model trained in stages that parallel developmental ones and that simulates cortical processes of self-organization during an early crucial stage of first word learning. This is done by taking into account strictly visual and acoustic perceptions only. The results obtained show evidence of the emergence in the artificial maps used in the model, of cortical functions similar to those found in the biological correlates in the brain. Evidence of non-catastrophic fast mapping based on the quantity of objects and labels gradually learned by the model is also found. We interpret these results as meaning that early stages of first word learning may be explained by strictly perceptual learning processes, coupled with cortical processes of self-organization and of fast mapping. Specialized word-learning mechanisms thus need not be invoked, at least not at an early word-learning stage.

Distributed intentionality: A model of intentional behavior in humans

Is human behavior, and more specifically linguistic behavior, intentional? Some scholars have proposed that action is driven in a top-down manner by one single intention—i.e., one single conscious goal. Others have argued that actions are mostly non-intentional, insofar as often the single goal driving an action is not consciously represented. We intend to claim that both alternatives are unsatisfactory; more specifically, we claim that actions are intentional, but intentionality is distributed across complex goal-directed representations of action, rather than concentrated in single intentions driving action in a top-down manner. These complex representations encompass a multiplicity of goals, together with other components which are not goals themselves, and are the result of a largely automatic dynamic of activation; such an automatic processing, however, does not preclude the involvement of conscious attention, shifting from one component to the other of the overall goal-directed representation. Marco Mazzone is Associate Professor of philosophy of language at the University of Catania. Emanuela Campisi is Visiting Researcher at the Max Planck Institute for Psycholinguistics in Nijmegen. Emanuela Campisi is mainly responsible for section 2, Marco Mazzone for the other sections. The authors, however, have discussed and planned together the whole paper, which is a prosecution of the line of research begun with Mazzone and Campisi (2010).

Constructing the context through goals and schemata: top-down processes in comprehension and beyond

My main purpose here is to provide an account of context selection in utterance understanding in terms of the role played by schemata and goals in top-down processing. The general idea is that information is organized hierarchically, with items iteratively organized in chunks—here called “schemata”—at multiple levels, so that the activation of any items spreads to schemata that are the most accessible due to previous experience. The activation of a schema, in turn, activates its other components, so as to predict a likely context for the original item. Since each input activates its own schemata, conflicting schemata compete with (and inhibit) each other, while multiple activations of a schema raise its likelihood to win the competition. There is therefore a double movement—with bottom-up activation of schemata enabling top-down prediction of other contextual components—triggered by multiple sources. Another claim of the paper is that goals are represented by schemata placed at the highest-levels of the executive hierarchy, in accordance with Fuster’s model of the brain as a hierarchically organized perception-action cycle. This account can be considered, in part at least, a development of ideas contained in Relevance Theory, though it may imply that some other claims of the theory are in need of revision. Therefore, a secondary purpose of the paper is a contribution to the analysis of that theory.

The continuum problem: Modified Occam's Razor and conventionalisation of meaning

According to Grice’s “Modified Occam’s Razor”, in case of uncertainty between the implicature account and the polysemy account of word uses it is parsimonious to opt for the former. However, it is widely agreed that uses can be partially conventionalised by repetition. This fact, I argue, raises a serious problem for MOR as a methodological principle, but also for the substantial notion of implicature in lexical pragmatics. In order to overcome these problems, I propose to reinterpret implicatures in terms of implicature-like effects delivered by non-inferential processes.

Automatic and Controlled Processes in Pragmatics

In utterance understanding, both personal and sub-personal aspects appear to be involved. Relevance theory (starting from Sperber and Wilson 1986/1995) and Recanati (2004) have respectively explored two alternative ways to conceive of those aspects and their interaction. Here a third account is proposed, in the light of the automatic-controlled distinction in psychology, and of recent views concerning the cooperation between these two modes of processing. Compared to Recanati (2004), the account proposed here assigns a larger role to automatic, associative processes; at the same time, it rejects the view that consciousness applies only to what Recanati calls secondary pragmatic processes. Consciousness is rather held to cooperate with associative processes in any aspect of pragmatic processing, irrespective of the pragmatic distinction between explicatures and implicatures. On the other hand, a close consideration of how associative and conscious processes plausibly interact makes it appear unnecessary the hypothesis of a specialized process for utterance understanding—such as the automatic, inferential mechanism put forth by Relevance theory.

In Learning Nouns and Adjectives Remembering Matters: A Cortical Model

The approach used and discussed here is one that simulates early lexical acquisition from a neural point of view. We use a hierarchy of artificial cortical maps that builds and develops models of artificial learners that are subsequently trained to recognize objects, their names, and then the adjectives pertaining to their color. Results of the model can explain what has emerged in a series of developmental research studies in early language acquisition, and can account for the different developmental patterns followed by children in acquiring nouns and adjectives, by perceptually driven associational learning processes at the synaptic level.

Artificial learners of objects and names

Naming requires recognition. Recognition requires the ability to categorize objects and events. What mechanisms in the brain underlie the unfolding of these capacities? In this article, we describe a neural network model in which artificially created individuals are exposed to visual stimuli and vocal sounds and are tested in experiments like human children. The model simulates, in a biologically plausible way, the process by which infants learn how to recognize objects and words through experience.

Neural plasticity and concepts ontogeny

Neural plasticity has been invoked as a powerful argument against nativism. However, there is a line of argument, which is well exemplified by Pinker (The blank slate: the modern denial of human nature, Penguin, New York, 2002) and more recently by Laurence and Margolis (in: Laurence and Margolis (eds) The conceptual mind: new directions in the study of concepts, MIT, Cambridge, 2015) with respect to concept nativism, according to which even extreme cases of plasticity show important innate constraints, so that one should rather speak of “constrained plasticity”. According to this view, cortical areas are not really equipotential, they perform instead different kinds of computation, follow essentially different learning rules, or have a fixed internal structure acting as a filter for specific categories of inputs. We intend to analyze this argument, in the light of a review of current neuroscientific literature on plasticity. Our conclusion is that Laurence and Margolis are right in their appeal to innate constraints on connectivity—a thesis that is nowadays welcome to both nativists (Mahon and Caramazza in Trends Cogn Sci 15:97–103, 2011) and non-nativists (Pulvermüller et al. in Biol Cybern 108:573–593, 2014)—but there is little support for their claim of further innate differentiation between and within cortical areas. As we will show, there is instead strong evidence that the cortex is characterized by the indefinite repetition of substantially identical computational units, giving rise in any of its portions to Hebbian, input-dependent plasticity. Although this is entirely compatible with the existence of innate constraints on the brain’s connectivity, the cerebral cortex architecture based on a multiplicity of maps correlating with one another has important computational consequences, a point that has been underestimated by traditional connectionist approaches.

What Kind of Associative and Inferential Processes?: A Response to Rubio-Fernández (2013)

Rubio-Fernandez (2013) is a noteworthy instance of the recent efforts of grounding speculative pragmatic models into empirical research. However, it also shows that between pragmatic theories and the relevant psycholinguist research there are still conceptual gaps. Specifically, Rubio-Fernandez reports two studies concerning emergent properties, that is, properties that are not associatively activated by related concepts but need to be inferred thanks to contextual information. These studies seem to show that emergent properties are activated by an early integration of associative and inferential processes and this is taken as evidence in favour of one-stage over two-stage models of pragmatic processing. I analyse the relevant notions of associative and inferential process in terms of, respectively, automatic activation of directly accessible information and consciously sustained activation of indirectly accessible information. These are not, however, the notions adopted in the pragmatic models considered by Rubio-Fernandez. At a closer analysis, therefore, none of these models is fully compatible with her evidence. My line of argument is based on a crucial, but too often ignored, distinction between two meanings of “associative”: as a behavioural description of cognitive processes and as a model for their implementation. I also illustrate some consequences of neglecting this distinction.