Concepts and Methods of Cognitive Semiotics
Göran Sonesson, Jordan Zlatev, Joel Parthemore
We investigate similarities and differences between ideas from phenomenology, semiotics, cognitive science and linguistics concerning basic concepts such as "sign", "representation", "meaning", "body", "intersubjectivity", and aim at a coherent synthetic theoretical framework for Cognitive Semiotics.
Methodologically, we explore the relation between First-person ("subjective"), Second-person ("intersubjective") and Third-person ("objective") methods, emphasizing the primacy of the first two types. Observational methods, such as those that we employ in the various empirical studies carried out by our group (such as those mentioned below) are indispensible, but they presuppose the analysis of (linguistic) intuitions and empathy ("others are basically like us") (cf. Sonesson 2007; Zlatev 2008; Itkonen 2008).
The Nature of Grammar: Between Convention and Motivation
Jordan Zlatev, Arthur Holmer
A cognitive semiotic perspective on grammar rejects the notion of an "autonomous" form, determined by principles independent of meaning and use. If (a specific) language is "a conventional-normative semiotic system for communication and thought" (Zlatev 2008), then grammar is a (specific) complex conventional pairing of meaning and expression (cf. Langacker 1987).
Yet, it remains an open question how from such a view to account for typological "universals", or at least tendencies (in word order, word-classes, alignment etc). A key to addressing this issue is that conventionality does not imply "arbitratiness", but simply common knowledge. Motivation, from semiotic, ecological, cognitive, functional and other possible factors can be seen as the source of contraints behind
grammatical variation. The nature of these "factors", however, remians unclear (cf. Croft 2003; Deacon 2003).
In this subproject, we investigate the interplay between convention and (different forms of) motivation for grammatical structure from a theoretical and empicial perspective.
Prosody, Information Structure and Word Order
Arthur Holmer and Anastasia Karlsson
It is generally assumed that some intonational patterns are initially physiologically determined (e.g. the drop of F0 at the end of a message as the logical consequence of declination, i.e. the gradual lowering of pitch during an utterance). It is further assumed that this pattern is then phonologized during language evolution. One of the questions to be addressed is whether the phonologization of tonal phenomena such as this is connected to or can be affected by word order and information structure. Here we will concentrate on the functional distribution of low and high tones and its possible connection to word order type: therefore, one of the initial stages of our investigation is the analysis of intonation in a typologically diverse sample of languages. Languages we intend to investigate in this respect are Kammu (SVO), Seediq (VOS), Bunun (VSO) and Puyuma (alternating VOS/VSO). As a second stage we plan if possible to extend our study to Japanese (SOV), Thai (SVO), Mongolian (SOV), Basque (free, predominantly SOV), Scottish Gaelic (VSO).
Our study concerns the connection between three related features in language: information structure, word order and prosody. The underlying assumption is that these three phenomena are intimately inter-related, and our goal is to identify how this interaction works. Two of the pairs involved in this interaction are already quite uncontroversial: it is already well-known that intonation can be used to convey information structure (focus intonation is an example of this), and it is equally well-known that information structure is relevant for word order (topicalization and scrambling are examples of this). The third pair, the connection between word order and intonation, has received very little attention and is as yet poorly understood. It is in this pair where we expect to make the most important typological contribution.
The first issue to be addressed concerns identifying intonational correlates of word order: if we generalize across languages with different word order patterns, do we find these orders reflected by intonational patterns? Can we use prosodic phrasing to identify syntactic constituents? A special test case concerns languages where word order varies between two alternative patterns: here we would ideally expect to find systematical prosodic reflexes of the different word order patterns. If such reflexes are found, the next step is to investigate whether these in any way correspond to the intonational patterns occurring in languages where the relevant orders are basic orders. The language we intend to investigate here is Puyuma (VOS/VSO), to look for shared patterns with Seediq (VOS) and Bunun (VSO) respectively. The relevant field studies have been conducted, and analysis is under way.
An intersemiotic perspective on narrativity
(Göran Sonesson, Anna Cabak Redei, Gunnar Sandin, Sara Lenninger, Michael Ranta, Erik Rynell)
Within the humanities, narratology has been a standing research area during the last 50 years, most notably among literary analysts, linguists, and semioticians. Quite frequently, narration has been associated with verbal discourses, whether written or oral, where, briefly put, events or situations are temporally ordered. Accordingly, theoretical discussions concerning narrativity have usually focused upon literature and drama, though also on film and television. Pictorial signs, gestures and other non-verbal semiotic resources have received comparably little attention. In this project, however, we intend to extend and deepen previous research within narratology in accordance with the general intersemiotic approach outlined above.
There are particular reasons for taking up narratives in relation to the study of human specificity, notably as it emerges in ontogeny, phylogeny, and history. Within cognitive science, narrative-like structures have figured prominently in two ways. Inspired by the memory schemas of Bartlett , early cognitive science introduced terms such as frames, scripts or event schemas [e.g. Schank & Abelson 1977; Mandler 1984] for actions and events belonging to more general categories. Through previous experiences we acquire a large amount of culturally based event and scene stereotypes (along with idiosyncratic variations), either due to direct familiarity with instances of events, or due to our acquaintance with written, oral, and of course pictorial descriptions of them (e.g. religious or mythological tales). However, scripts primarily intervene at action levels as such, not specifically at those of narrated action. Donald (1991) thus distinguishes in his phylogenetic approach mimetic actions from the much later “mythic” stage, where stories begin to be told.
In more recent cognitive science, narrativity has come to play a rather different part, as a constituent of evolution and/or child development. In Donald’s model, narrativity only emerges at the 4th stage, just before leaving biological evolution behind and entering cultural history, and at the same time as language. In her parallel account of child development, Nelson (1997; 2007) also identifies language and narrative, presenting the latter as a very important factor in the evolving mind. Moreover, narratives (esp. autobiographical ones) have also been argued to play a significant role as to the establishment of human personal identity and indeed consciousness itself [cf. Bruner 1990; Neisser & Fivush 1994; Fireman et al. 2003] as well as in the constitution of group identities [cf. Bartlett 1932/1995; Bruner 1990; Donald 1991; Schank 1995; Pennebaker et al. 1997; Hutto 2008]. These ideas go together with the identification of narrative and language (Donald 1991; Hutto 2008; Fireman et al. 2003), which both in the end are supposed to stand in for mental life. In the extreme case, narrative, identified with language, as well as with the self, is taken to be the only thing in the human mind which is not simply “subpersonal”, that is, functioning automatically and unconsciously (Dennet 1992; Gazzaniga 1998). If narrativity is not merely verbal, considerable doubt is thrown on these ideas.Classical narratology may offer some tools for sorting out these issues. Although the study of literature continues to dominate the field, other semiotic resources have of course been studied, for example within art history [e.g. Panofsky 1939 (1962); Brilliant 1984; Lavin Aronberg 1990, Bal 1991; Greenstein 1992; Derbes 1994; Andrews 1995]. Still, attempts to elucidate the cognitive fundaments of visual narrativity, esp. in static pictures, have scarcely occurred [some exceptions being Theissing 1987; Pochat 1996; Sonesson 1997; Nanay 2009; Ranta 2001/02, 2007, 2010 a/b]. Narrativity in film and television has, for obvious reasons, received comparably more attention [e.g. Bordwell 1985; Kozloff 1988; Jost 1989; Chatman 1990; Deleyto 1996 ; Branigan 1992].
The exact nature of what constitutes narratives has been and still is a matter of standing dispute within contemporary research. A minimal requirement for something being a narrative has, for example, been claimed to be the representation of “at least two events with a temporal link on the content side.” (Prince 1982). No particular requirement thus seems to be imposed on the expression side [cf. Sonesson 1997], which opens up for the possibility of narrative being enacted in media other than language. It is easy to find objections to the classical definition of narratives formulated by e.g. Prince. Thus, “The child cried. Its mother picked it up” may well be a story, but it is not a particularly “good” (prototypical) one. Indeed, Prince (1996), Coste (1989), and Ryan (1991) have suggested various criteria for something being a good story, and it was suggested by Sonesson (1997) that some of these criteria fit perfectly also with static pictures, while others do not. Some of these criteria may however turn out to characterize actions rather than only their representation.
Thus we may conceive of narratives as - to various degrees - representations of (i) sequential, causally related orders of events; (ii) particular rather than general events/states; (iii) intentional states (e.g. beliefs, desires); (iv) elements of conflict between different subjects; (v) virtuality or disnarrated elements (implicit/explicit alternative courses of action); (vi) hermeneutic composability, i.e. the interpretation of individual occurrences interacts with that of larger plots or contexts - just to mention some examples (cf. Bruner 1991).
Using these and other criteria for more or less prototypical narrativity, we will investigate whether narrativity is uniquely or predominantly verbal, as opposed to being also nonverbally manifested (e.g. in drama, painting, film etc.). Given the many complex theoretical issues involves, as illustrated above, a large part of the project must no doubt be dedicated to theoretical analysis, both of the notion of narrativity, and to its application to different semiotic resources. Another considerable part, however, will consist of empirical studies attempting to corroborate our intuitions about non-verbal narrativity. Even Prince (1999) has quite recently pointed to the necessity of empirical studies, but hardly any seem to have been done so far.. The investigations within our project will be carried out as follows:
1) Narrativity in static pictures and picture sequences (including film).
Pictorial stimuli will be varied systematically, to investigate whether they manifest one or several of the criteria of narrativity. The subjects will be adults, but children will be added later on.
2) Narrativity in action
Bodily movements will be varied systematically, studied in relation to different media, and evaluated according to narrative criteria.
To summarize, , while considering the significance of narrative in constituting human identity, the focus of our research will fall upon the following research issues:
1. Narrativity in non-verbal as well as verbal semiotic means
2. Relations between the organization of action and narratives.