In the previous sections, I have sketched an overall characterization of the conceptualizer and the dimensions of construal, while emphasizing certain points that are relevant for the up-coming discussion. Instead of describing these analytical concepts in full, I have aimed to illustrate their interconnectedness and their applicability to analysis of actual (in this case, written) discourse. The interconnectedness of the dimensions of construal has been amply discussed in previous literature. For example, Verhagen (2007:
53–58) discusses the different classifications for construal in the Cognitive Linguistic literature. These classifications, as I have mentioned, vary even within Cognitive Grammar, and other authors, such as Talmy (1988, 2000a), as well as Croft and Cruse (2004), have presented their own taxonomies of construal phenomena.
Croft and Cruse, who aim to synthesize Talmy and Langacker’s classification systems, point also to the inevitable amount of arbitrariness in any given segmentation of construal phenomena (Croft & Cruse 2004: 43–46).
Verhagen (2007), on the other hand, points out how the increase in linguistic coverage attained by Croft and Cruse’s system seem to increase also its arbitrariness, making it difficult to see how the posited categories are
motivated by underlying conceptual operations; motivation which nevertheless is necessitated by the cognitive commitment.
Verhagen (2007: 57) brings this difficulty back to the double nature of construal: its structure is based on putative conceptual operations, but it actually classifies linguistic units, which are in constant change. It is thus expected that a certain structure may correspond to two different construals either for a language-using subject or within speech community. This makes a strict categorization of construal not impossible but cognitively unrealistic.
This obviously does not concern the speaker, who uses and mentally processes the structure in question according to the conventional linguistic knowledge she possesses.
At the same time, the speaker’s activity involves simultaneously multiple facets of processing that correspond to different dimensions of construal, which requires that any comprehensive (or so intended) analysis of an utterance must involve multiple or all dimensions of construal. While this task is possible, it is rather unrealistic to assume that a final conclusion may ever be achieved even for a specific utterance or that the analysis is able to pinpoint the relations between different dimensions of construal in a way that is ontologically realistic. Rather, the dimensions of construal are discernible exactly as analytical tools, whereas the facets of meaning they capture are fluid, evasive and dependent of the conceptualization they characterize. It is not therefore unexpected that Langacker himself states that the segmentation of construal into separate dimensions is “mostly for expository convenience”
(Langacker 2007: 421–462, n. 22).
The other characteristic of construal that I have illustrated is its double nature. Construal expresses the subjectivity of the conceptualizer, and this subjectivity corresponds to alternative construals for the same objective situation. (In some cases, such as in that of profiling, construing the same conceptual content may lead to referential difference). These aspects of construal, subjectivity and selection, may and often do result as semantic patterns within discourse. In this sense, construal is not only a theoretical relationship between an actual expression and its potential alternative (the paradigmatic plane), but an actual interrelation between semantically related expressions within a written or spoken discourse (the syntagmatic plane).
In both of these cases, the subjective aspect of construal can be easily over-emphasized. The construal phenomena discussed above may be categorized as subjective in the sense of ‘non-objective’, for they inevitably evoke the conceptualizing subject and her perspective. However, as noted in Verhagen’s (2005, 2007) analysis of the ground, the conceptualizer’s perspective is in many (if not most) cases construed as being in alignment with the perspective of another conceptualizer. On an even more intersubjective level, the evoked conceptualizer of an expression is not simply a conceptual origo, but also a perspectival position the recipient is invited to assume. In this sense, construal is reactive to the interactional context and thus a function of the ongoing discourse. As the discourse involves multiple participants by default, the very
premise of positing construal as a (primarily) conceptual phenomenon can be questioned. The very possibility of such questioning demands a closer look at the theoretical groundwork of Cognitive Grammar.
3 CONSTRUAL AND IMAGERY
This chapter describes in detail the theoretical foundations on the basis of which Cognitive Grammar argues for its conception of linguistic meaning. The chapter is primarily concerned with the argumentative relationship between Cognitive Grammar and the existing literature it refers to, rather than the linguistic theory as such. Regrettably, the reference made by Cognitive Grammar to its theoretical precursors is relatively scarce, which requires that any claims made about its theoretical argumentation are prefaced by thorough reconstruction. Therefore, to enable a detailed but concise analysis, I must concentrate on one specific aspect of Cognitive Grammar’s argumentation: its use of the notion of “imagery”.
By imagery, Cognitive Grammar refers both to linguistic and mental types of iconic representation (see section 2.2.1 for discussion on iconicity in Cognitive Grammar). The relationship between these two variants proves essential for understanding the way in which Cognitive Grammar conflates distinct ontological levels in its attempt to ground meaning in mental activity. Thus, the question of imagery can serve as an illustration of Cognitive Grammar’s ontological stance as a whole. Furthermore, the term ‘imagery’ is used in tandem with construal to refer to the non-objective side of linguistic meaning, and is therefore directly relevant for this treatment of the dimensions of construal.
Indeed, imagery ties together the argumentative structure and the analytical apparatus of Cognitive Grammar.
Neither the significance of imagery in Cognitive Grammar nor the associated argumentative shortcomings of the theory have been left unnoticed in previous literature. My analysis will discuss the criticisms levelled against Cognitive Grammar to the effect that the dimensions of construal are based on a mentalist conception of meaning (e.g. Itkonen 1997, forthcoming), i.e. that they compose meaning as it is represented in the mind of an abstracted psychological subject, and that this mentalistic conception of meaning is physicalist/reductionist by nature (Kenttä 2003).
I accept both criticisms to the extent that Cognitive Grammar suffers from inadequate argumentation and ontological inconsistencies. I also concur that there exists a strong justification for the conclusions to which both lines of criticism lead. Cognitive Grammar does indeed explicitly state that meaning lies in the head of a single representative speaker/hearer, and it therefore conflates two ontological levels, that of higher mental activity (conceptualization) and that of its physical basis (neural activation). However, both lines of criticism fail to recognize characteristics of Cognitive Grammar that speak against these conclusions. First of all, Cognitive Grammar is anything but definitive concerning ontology and the relationship between the individual and the social level. Second, while Cognitive Grammar emphasizes mental structure in the description of meaning, it nevertheless is a usage-based model that continuously refers to the social aspect of language acquisition and usage; and, as a result from its
non-modular and experiential view of cognition, Cognitive Grammar is more heavily based on social phenomena than it is willing to admit.
Together these reservations make another constructive line of criticism possible. This criticism considers Cognitive Grammar a linguistic theory compatible with a social account of meaning and communication. In so being, the approach allows for a better appreciation of the argumentative structure between Cognitive Grammar and its theoretical sources of influence. This, however, requires a more subtle understanding of the social level – both in general and relative to the formation of subjectivity – than is made possible by Itkonen’s insistence of the strong autonomy of linguistics or by Kenttä’s focus on the argumentative weak-points in Cognitive Grammar.
Regardless of the psychologistic proclamations it puts forward, there is no doubt that Cognitive Grammar is more about ‘grammar’ than it is about
‘cognitive’ aspects of language, and this is also evident in its theoretical prerequisites. This is not to say that Cognitive Grammar’s ontological delineation of language as a cognitive phenomenon does not bear effect on how language is segmented into substructures and properties. Rather, this ontological mentalism calls for close scrutiny on how Cognitive Grammar deals with the divide between the individual and the social aspects of conceptual meaning. The focusing on imagery in this chapter serves exactly this goal. By concentrating on the dual nature of the notion of concept and the semantic properties that are categorized by this notion, I will show that the metatheoretical self-conception of Cognitive Grammar is distorted by a vague conception of what counts as ‘cognitive’.
In this respect, my analysis is in line with the tenacious criticism of Cognitive Grammar by Itkonen. But if there is something fundamentally wrong with the general orientation of basing linguistic meaning on mental processes, then why analyze the details of the theory’s reference to some psychological studies instead of simply stating the invalidity of such a practice? The argument I make in this chapter is that the individual-psychological interpretation of these sources that Cognitive Grammar provides is not the only one possible, and that the way Cognitive Grammar describes linguistic meaning may be informative with regard to the necessary interaction between individual and the social level of linguistic organization. The analysis at hand will therefore be begun by stating the question that is central to the present concerns: how semantics as envisaged by Cognitive Grammar should be positionedvis-à-vis psychological and social ontology.
3.1 OUTLINE
The object of analysis, i.e. the argumentation for a specific semantic conception of Cognitive Grammar, is approached here with a focus on how it justifies, or fails to justify, the ontological stance assumed by the theory, on the one hand, and the structuring of meaning into the triad of conceptualization, conceptualizer, and construal, on the other. The treatment of these questions will emphasize the role of construal, i.e. the organization of semantic content. The metatheoretical analysis of semantic theory will hence apply to the dimensions of construal
specifically. The two main sections of this chapter correspond to two standpoints from which the theoretical structure of Cognitive Grammar may be approached:
section 3.2 discusses the essential features of Cognitive Grammar’s theoretical self-justification from a general ontological perspective, whereas section 3.3 focuses to the notion of imagery as a particularly illustrative facet or this justification. The summary in section 3.4 provides a synthesis of the argumentative structure revealed by the preceding chapters’ analyses and considers its significance for the concept of construal.
On occasion, the following analysis will delve into minute details of Cognitive Grammar’s theoretization on linguistic meaning. While not at all historical, the present analysis will also address the changes that occurred in the organization of the dimensions, when these prove to be theoretically significant. The data for the analysis has been selected from the volumes FCG-I and CGBI, which cover over two decades of the theory’s development. I will continue to use the term
“dimensions of construal”, but when there is the need to emphasize some specific phase in the development of this notion, I will use the term “focal adjustments”
(e.g. FCG-I) or “dimensions of imagery” (e.g. CIS).