VIEWPOINT IN LINGUISTIC DISCOURSE
Space and evaluation in news reports of political protests
Christopher Hart
This paper continues to develop a programme of research which has recently emerged investigating the ideological functions of spatial construals in social and political discourse from a Cognitive Linguistic perspective. Specifically, inspired by principles in Cognitive Grammar, the paper attempts to formulate a grammar of āpoint of viewā and show how this trans-modal cognitive system is manifested in the meanings of individual grammatical constructions which, when selected in discourse, yield mental representations whose spatial properties invite ideological evaluations. The link between spatial organisation and ideological evaluation in these mental models, it is argued, is a function of our embodied understanding of language. These theoretical arguments are illustrated with data taken from online news reports of two political protests.
1. Introduction
In critical discourse analysis (CDA) there is an increasing focus on the visual articulation of discourses and thus the parameters involved in the visual expression of ideology (e.g. Abousnnouga & Machin, 2011; Machin & Mayr, 2012). Researchers in multimodal CDA see linguistic and visual modes of communication as manifestations of a single underlying semiotic capacity. For Kress and Van Leeuwen, for example, meaning belongs to culture and, although realised quite differently, many of the same meaning potentials may find parallel expression across alternative modalities (2006, p. 2). In developing a grammar of visual design, multimodal researchers have therefore drawn on categories that make up functional linguistic grammars of the kind developed and applied in Critical Linguistics. Thus, multimodal researchers have studied visual instantiations of grammatical systems like TRANSITIVITY and MODALITY (Kress & Van Leeuwen, 2006; Machin, 2007). Visual instantiations of conceptual metaphors have also been analysed (Bounegru & Forceville, 2011; El Refaie, 2003). The relationship between linguistic and multimodal approaches to CDA, however, has tended to be one of unidirectional influence where linguistic approaches can be seen to inform visual approaches but the converse is not observed. For example, an elaborated grammar of visual design takes in a number of additional systems which are assumed to be unique to visual communication and therefore do not typically feature in the analyses of linguistic approaches to CDA. Amongst such systems is spatial POINT OF VIEW (PoV).
In ordinary parlance, of course, āpoint of viewā is taken to mean something akin to ideology or evaluation. Point of view in this sense has been much studied in Critical Linguistics where it is said to be reflected in the āchoicesā a text presents in grammatical systems like TRANSITIVITY and MODALITY (Fowler, 1991; Hodge & Kress, 1993; Lee, 1992; Simpson, 1993). Space and spatial PoV, however, have not figured in any systematic way in linguistic approaches to CDA (cf. Chilton, 2004). This is in spite of: (i) relevant findings from Cognitive Linguistics which highlight the role of spatial cognition and thus the presence of spatial patterning in linguistic meaning (Lakoff & Johnson, 1980; Langacker, 1987, 1991; Talmy, 2000) and (ii) as has been demonstrated in multimodal CDA, the potential for alternative spatial configurations and PoVs to communicate ideology and function rhetorically (Kress & Van Leeuwen, 2006).
According to Cognitive Linguistics, as well as Cognitive Science more generally, language is embodied (Croft & Cruse, 2004; Johnson, 1987; Lakoff & Johnson, 1999). To say that language is embodied is to say that it is grounded in physical, including visuospatial, experience and is bound with the bodyās situatedness at the point of interpretation. This means at least three things:
ā¢that language relies on systems whose primary role is in support of other non-linguistic domains like vision and action. In other words, that the conceptual processes involved in language and discourse are not principally distinct from processes that function in other experiential realms like visual perception and spatial cognition. Consequently, linguistic understanding will in many cases have correlates in our experiences of other kinds of meaningful episode;
ā¢that much of meaning is provided by conceptual structures which develop pre-linguistically based on the kind of bodies we have, our interactions with the physical environment, and co-occurrences of physical and mental experiences. Such conceptual structures include image schemas and conceptual metaphors;
ā¢that language is necessarily interpreted with reference to the Self, including the location and orientation of the Self in space and time. This can be seen most clearly in the phenomenon of deixis.
One upshot of language being embodied is that many linguistic (lexical, grammatical, and textual) constructions include, as part of the meanings they invoke, properties related to visuo-spatial experience. For example, Langacker (2008, p. 75) argues that many, if not all, grammatical constructions invite the conceptualiser to construe the scene described from a particular PoV. Talmy (2000) similarly shows that grammatical constructions confer a figure/ground relation on elements in the scene described. In a further development of this framework, it has been argued that meaning involves running a mental simulation of the event described seen from the perspective of an āimmersed experiencerā (Bergen, 2012; Zwaan, 2004). That is, we āseeā (imagine) the events described in our mindās eye in a way that is similar to a multimodal semiotic experience. These programmes both point to the role of mental models in discourse processing and can thus usefully add to the van Dijkian notion of an āevent modelā (e.g. van Dijk, 1997, 1999, 2008). They view language as a set of prompts which, drawing on prior experiences, serve to guide the hearer in constructing an intersubjective mental representation of the referential situation. It is this mental model that forms the basis of our understanding of any utterance. For van Dijk, event models are made up of at least two components: a semantic or representational component and an affective component. The semantic component is assumed, ālacking alternative formats of representationā (1997, p. 191), to be propositional in nature. Drawing on insights from Cognitive Grammar and Simulation Semantics, however, we can begin to think more precisely about the form that these models might take. On the account presented here, the mental models built in discourse are, at one level of representation, schematic consisting of an abstract conceptual structure. At another, they are more fully specified presenting a rich and dynamic simulation of the event unfolding in (imagined) space and time.These representations are then coupled at a third level with epistemic, moral and affective evaluations (which we may begin to think about in the terms laid down in Martin and Whiteās (2007) grammar of Appraisal (see Hart, 2014)).
One particularly fruitful avenue for CDA to pursue, then, might be to investigate the role and effects of embodied spatial construals, and PoV operations in particular, in social and political linguistic discourse. More recently, CDA has witnessed a cognitive turn (see Hart, in press for an overview) in which researchers, drawing heavily on insights from Cognitive Linguistics, have indeed become more interested in the role of space and spatial concepts in structuring social and political domains of experience (Cap, 2013; Chilton, 2004; Dunmire, 2011; Filardo-Llamas, 2013; Hart, 2013a, 2013b, 2014; Kaal, 2012). For example, in Critical Metaphor Studies it has been shown that abstract notions like the nation state, economic and social conditions, and political progress get structured metaphorically in terms of more familiar concepts like CONTAINMENT, VERTICALITY, and FORWARD MOTION, which are all closely linked with experiences we have of our bodyās existence in space (Beer & De Landtsheer, 2004; Charteris-Black, 2004; Chilton, 1996; Hart, 2010).
In a distinct but closely related programme of research known as Discourse Space Theory (DST), Chilton (2004) suggests that meaning construction involves the metaphorical āmapping outā of discourse elements within a particular type of mental space (Fauconnier, 1994, 1997). According to this model, people, places, attributes, and events introduced in discourse get located along three axes ā space, time, and modality ā at relative distances from a deictically specified anchorage point in the centre of an abstract three-dimensional discourse space. Actions are represented in the model as abstract metaphorical movements through the space represented by elements which can be formally defined as vectors.1 One crucial feature of this framework, then, is that it suggests a geometric mode of conceptualisation in which spatial concepts play a fundamental structuring role. Another is that it argues for the ego-centric nature of meaning construction. The deictic centre represents the conceptualiserās āsituatednessā within the abstract space and thus the PoV from which they construe the world presumed by the text. The deictic centre is decoupled from immediate situational context, however, to represent what the conceptualiser takes more broadly as their social, temporal, epistemic, and deontic āgroundā (see Hart, 2014). Mentally situated at the centre of the discourse space, the hearer is placed āonstageā as an object of conceptualisation and thus themselves form an integral part of the meaning of any text. In a further elaboration of this framework, Cap (2006, 2013) identifies as a feature of interventionist discourse a rhetorical strategy of proximisation. In its spatial dimension, proximisation involves a construal of discourse participants defined (explicitly or implicitly) as alien or antagonistic, and thus positioned initially as distal along the spatial axis, encroaching on or entering the conceptualiserās territory at deictic centre. Such a conceptualisation serves to construct threat-salience on the back of which interventionist action may be legitimated. Proximisation may be particularly effective in evoking feelings of anxiety. When the conceptualiser or their spatial ground is construed as the target of the proximisation, the conceptual process would have as a correlate in visuo-spatial experience that of an unfamiliar or unfriendly entity entering oneās peripersonal space.
DST postulates a ...