Neuroscience and Biobehavioral Reviews 69 (2016) 69–78 Contents lists available at ScienceDirect Neuroscience and Biobehavioral Reviews jou rnal homepage: www.elsevier.com/locate/neubiorev Grounding meaning in experience: A broad perspective on embodied language a,∗ b c d Giovanni Buccino , Ivan Colagè , Nicola Gobbi , Giorgio Bonaccorso a Dipartimento di Scienze Mediche e Chirurgiche, University “Magna Graecia” of Catanzaro, Viale Salvatore Venuta, 88100, Germaneto, Italy b Faculty of Philosophy, Pontifical University Antonianum, Via Merulana 124/B, 00185, Rome, Italy c Faculty of Philosophy, Pontifical Gregorian University, Piazza della Pilotta 4, 00187, Rome, Italy d Institute for Pastoral Liturgy of Padua, Affiliated to the Pontifical Athenaeum St. Anselmo in Rome, Via G. Ferrari, 2/A, 35123 Padua, Italy a r a t i c l e i n f o b s t r a c t Article history: This work reviews key behavioural, neurophysiological and neuroimaging data on the neural substrates Received 1 March 2016 for processing the meaning of linguistic material, and tries to articulate the picture emerging from those Received in revised form 29 May 2016 findings with the notion of meaning coming from specific approaches in philosophy of language (the Accepted 27 July 2016 “internalist” view) and linguistics (words point at experiential clusters). Available online 28 July 2016 The reviewed findings provide evidence in favour of a causal role of brain neural structures responsible for sensory, motor and even emotional experiences in attributing meaning to words expressing those Keywords: experiences and, consequently, lend substantial support to an embodied and “internalist” conception Embodiment Internalism of linguistic meaning. Key evidence concern verbs, nouns and adjectives with a concrete content, but the challenge that abstract domains pose to the embodied approach to language is also discussed. This Philosophy of language Sensorimotor systems work finally suggests that the most fundamental role of embodiment might be that of establishing com- Mirror neuron system monalities among individual experiences of different members of a linguistic community, and that those Canonical neuron system experiences ground shared linguistic meanings. Semantics © 2016 Elsevier Ltd. All rights reserved. Contents 1. Introduction . 69 2. Meaning in philosophy of language . 70 3. Neuroscience, “embodied” language and meaning . 72 3.1. Understanding verbs . 72 3.2. Understanding nouns . 73 3.3. Meaning and experience . 73 4. Possible objections and perspectives . 74 4.1. Abstract linguistic domains . 74 4.2. The problem of solipsism . 75 5. Conclusion . 76 Acknowledgments . 77 References . 77 1. Introduction How do we understand the meaning of words, the basic ele- ments of our linguistic practice? Asking this question might seem ∗ odd, as we live immersed in a linguistic environment. However, Corresponding author. such a question has been extensively investigated since centuries. E-mail addresses: [email protected] (G. Buccino), Philosophy addressed this issue earlier than modern linguistics. [email protected] (I. Colagè), gobbi [email protected] (N. Gobbi), [email protected] (G. Bonaccorso). Nowadays, neuroscience brings key data into this inquiry, and time http://dx.doi.org/10.1016/j.neubiorev.2016.07.033 0149-7634/© 2016 Elsevier Ltd. All rights reserved. 70 G. Buccino et al. / Neuroscience and Biobehavioral Reviews 69 (2016) 69–78 seems ripe for articulating those data with other approaches to sion. The stereotype is a conventional, socially defined notion that language and linguistic meaning. gathers important and typical features of the extension of a word In this review, we will first summarize the main basic trends or expression. The extension is the set of items a word or expres- in philosophy of language about the issue of meaning. We will sion refers to. A speaker must know the socially defined stereotype start by considering how this problem was originally dealt with for acquiring the meaning. Moreover, under this perspective, the in early philosophy of language (Frege, 1892; Russell, 1905, 1910) extension defines the stereotype and, consequently, the meaning and then focusing on some late-twentieth-century developments of a word or expression. In Putnam’s view, this implies that the in philosophy (Putnam, 1975; Searle, 1969, 1979). Namely, we will meaning does not depend on what the speakers think, perceive, assess two fundamental streams in the philosophy of language, the experience etc., but on the “real nature”, the true (physical, chem- so-called “externalist” perspective and the “internalist” approach. ical, genetic, etc.) “hidden structure” of what the expression refers Does understanding the meaning of words resort to “external” to. This is so even when the speaker does not know such nature or (physical or social) entities or to “internal”, embodied experiences? hidden structure (Putnam, 1975 pp. 145-6 and 166). Therefore, the The central section of the review will focus on very recent extension determines the conventional/social stereotype, and the findings in cognitive neuroscience strongly suggesting that the stereotype is what speakers must grasp to understand meanings. meaning of several classes of linguistic items (i.e, verbs, nouns Despite differences between Frege’s and Putnam’s philosophy and adjectives), especially when indicating concrete actions, objects of language, the two positions share a rather “externalist” and dis- and intrinsic features of objects relevant for action, is grounded embodied conception of meaning. According to this conception: (a) in sensorimotor experience and related brain circuitries. The main meanings are supposed to be understood by grasping some kinds idea emerging from these findings, in keeping with the embodied of (physical or social) external entities (like senses or stereotypes), approach to language, is that the speakers understand linguistic and (b) “the psychological state of the individual speaker does not material thanks to a reactivation of those sensorimotor systems determine what he means” (Putnam, 1975 p. 192). involved in the experiences expressed by that linguistic material. Soon after Frege’s proposal, Bertrand Russell (1905, 1910) Thus, we will discuss the aforementioned streams in the philosophy reacted against both the notion of sense and the general idea of language in the light of recent findings coming from neuro- that meanings are unrelated to real and concrete experiences. He science, and show how they offer evidence and arguments in favour claimed that we can understand only those expressions that are of the “internalist” approach. entirely composed of constituents we are acquainted with, where Then, we will briefly consider a recent proposal in linguistics “to be acquainted with . .” essentially means “to have been directly (Dor, 2015) that stresses the experiential ground of linguistic mean- presented with . .”, or simply “to have experienced . .”. It is impor- ing, i.e. the idea that words point at clusters of experiences the tant to note that, in Russell’s view, we are not directly acquainted subject has had with the objects, actions or situations language is with objects, but with “sense-data” coming from our experiences about. with objects (Russell, 1910). Consequently, for example, we can A following section will deal with some implications of the per- understand the word “unicorn”, though we certainly cannot be spective emerging from the previous sections, addressing the issues acquainted with unicorns, if two conditions hold: 1) if we know (or of (i) the meaning in more abstract linguistic domains, and (ii) the are told, or read in, e.g., a dictionary) that a unicorn is a horse with a inter-subjectivity of meaning. horn on the head, and 2) if we have experienced horses, heads and horns. Note that condition 2) should not be understood, from Rus- sell’s standpoint, as meaning that we know the “essence” of such 2. Meaning in philosophy of language objects like horses, heads and horns, but that we are acquainted with certain clusters of sense-data coming from our experiences It may seem obvious that the meaning of a word is its refer- with horses, horns and heads. Moreover, if we learn the word “uni- ence (or “extension”), i.e. the thing, or set of things indicated by corn” because somebody explained it to us verbally (or as we read the word. However, modern philosophy of language – born with its definition on a dictionary), our understanding of “unicorn” also Frege’s (1892) “Sense and Reference” (see also Dummett, 1973; implies that we have learnt to associate the words “horse”, “horn”, Colagè, 2013) – began questioning this point. Frege noted that two and “head” to certain clusters of sense-data. In this way, the defini- expressions with the same reference (e.g. “the morning star” and tion of unicorn as a horse with a horn on the head serves the task of “the evening star”, both denoting Venus) do not necessarily have linking together our past experiences of horses, horns, and heads. the same meaning. Indeed, compare “the morning star is the morn- This opens up a more “internalist” and embodied stance about ing star” with “the morning star is the evening star”: if meaning meaning, according to which the meaning of linguistic expressions were all about reference, these assertions should have the same is fundamentally rooted in our experiences. We will see in Section meaning, which is not the case as the first is a mere tautology 3 that such experiences can be understood