1. INTRODUCTION
Originating in concrete, everyday interactions, it can be argued that language is a fundamentally social and, moreover, an intersubjective entity (e.g. Itkonen Reference Itkonen1978, Reference Itkonen, Haukioja, Helasvuo and Miestamo1997, Reference Itkonen2008b; Schegloff Reference Schegloff, Bornstein and Bruner1996; Zlatev Reference Zlatev2008b). This dictum concerns both linguistic expressions and their conventional meanings alike. Context-dependent, pragmatic meanings or pragmatic aspects of meaning are obviously a default characteristic of almost any linguistic communication. Their graspability, however, does depend on the presence of context-independent meaning that has been established by previous linguistic acts. That is, linguistic expressions are understood according to the conventions or norms that have already been put in place and have characterized their correct use (Itkonen Reference Itkonen, Haukioja, Helasvuo and Miestamo1997, Reference Itkonen2008b).
At the same time, meaning conveyed by linguistic expressions is characteristically non-objective, by which I mean that the public object of meaning – the designatum – does not determine that meaning alone.Footnote 1 Rather, a linguistic expression always reflects a certain human perspective vis-à-vis the designatum by explicating its certain facets while dissipating others, and organizing the selected facets according to an accepted and conventional point of view (e.g. Langacker Reference Johnson1987:116–137; Reference Langacker2008:55–89). For example, a person can be denoted either as a man or a linguist, and the same situation or process may be portrayed, for instance, either as progress (e.g. The young man I once knew developed into a fine linguist) or regress (The young man that I once knew is now nothing more than a petty linguist). Thus, these two expressions may very well portray a single factual occurrence but they do so relative to the utterer's more or less subjective evaluation of the occurrence.
With regard to their ontological status, it follows that linguistic meanings are social entities, but simultaneously they are also non-objective or perspectival with regard to their expressive content. Thus, linguistic expressions convey conventionalized perspectives that manifest human subjectivity in an intersubjectively valid manner. One may, therefore, characterize linguistic meanings as socially constituted non-objective depictions of their designata.
Accordingly, a valid semantic theory should accommodate the social constitution as well as the non-objective quality of the linguistic meaning. The Cognitive Linguistic paradigm (e.g. Talmy Reference Stern2000; Verhagen Reference Trevarthen and Aitken.2005, Reference Verhagen2007, Reference Verhagen, Geeraerts and Cuyckens2008; Langacker Reference Langacker2008) scrutinizes non-objective meanings as an inherent property of language. Further, Cognitive Linguistics study has also recently started to show a growing interest in the social nature of linguistic meaning (e.g. Sinha Reference Schegloff, Ochs, Thompson and Schegloff1999; Zlatev Reference Zlatev2010; see also Etelämäki et al. Reference Etelämäki, Herlin, Jaakola and Visapää2009, Etelämäki & Jaakola Reference Etelämäki and Jaakola2009, Etelämäki & Visapää Reference Etelämäki and Visapää2014, Möttönen Reference Meltzoff and Keith Moore2016). What has been lacking in this undertaking, however, is a theoretical model that explicitly affirms and concentrates on the interdependence that exists between the social constitution of linguistic meaning and the non-objectivity of that linguistic meaning.
In this paper, that model is developed based on a phenomenological analysis of linguistic meaning (Banchetti-Robino Reference Banchetti-Robino1997; Zlatev Reference Zahavi2007a, b; Woelert Reference Verhagen2011; Blomberg & Zlatev Reference Blomberg and Zlatev2013). Phenomenology originated in the early 20th century in the work of Edmund Husserl (e.g. Husserl Reference Husserl2001a, b), who established the doctrine as a rigorous study of consciousness and its properties that then make coherent experiences and objective knowledge possible. I argue that the two key notions of Husserl's phenomenology, namely, intentionality and intersubjectivity, can be applied effectively to explain both non-objectivity and sociality characteristic to linguistic meaning. Specifically, I suggest that non-objectivity pertains to a conventionalized intentionality that exists between the subject and the object of a linguistic sign. Moreover, I argue that this relation presupposes pre-linguistic intersubjectivity, namely, the ability to experience foreign intentions and take part in the social constitution of objects as shared foci of multiple intentions.
I thus also argue that language manifests the very properties of consciousness that are the a priori conditions for linguistic capacity at large. These properties exist prior to language both ontologically and developmentally. In developmental psychology, the human capacity to recognize and identify with conspecifics prior to language skills is a well attested fact that scholars with different theoretical backgrounds do agree upon (see e.g. Trevarthen Reference Tomasello1979, Reference Trevarthen and Bullowa1980; Astington Reference Astington2006; Gallagher & Hutto Reference Gallagher and Hutto2008; de Bruin & de Haan Reference De Bruin and de Haan2012). This viewpoint has also been adopted as a necessary starting point for both theorization and analysis by many scholars of Conversation Analysis, Interactional Linguistics (e.g. Goodwin Reference Goodwin1981, Reference Goodwin2000, Reference Goodwin2006; Schegloff Reference Möttönen1989) and intersubjectively grounded theories of language development and evolution (e.g. Tomasello Reference Talmy2003, Zlatev Reference Zlatev2008b).
In contrast, there have been few studies that have directly focused on the implications of pre-linguistic intersubjectivity for representational semantic content or non-objective meaning in particular (see, however, Verhagen Reference Trevarthen and Aitken.2005; also Zlatev & Blomberg Reference Zlatev, Schmicking and Gallagher2016 in this volume). The model developed here focuses directly on this relationship, thus providing a conceptual basis for analyzing non-objective meaning as a linguistic manifestation of pre-linguistic intersubjectivity.
The task at hand then is pronouncedly theoretical. Consequently, the present treatise is concerned with the prerequisites of semantic analysis rather than particular empirical phenomena per se. The claims made regarding semantic analysis have, however, practical consequences for empirical semantic study, as will also be discussed. This aspect will be emphasized here by adopting a particular semantic theory, namely Langacker's (Reference Johnson1987, Reference Langacker2008) Cognitive Grammar, as the basis for a phenomenologically inspired model of non-objective meaning. In a recent treatment (Möttönen Reference Meltzoff and Keith Moore2016), I analyzed at length the ontological reductionism of Cognitive Grammar, while yet showing that the theory includes a solid descriptive apparatus for the precise analysis of non-objective meaning (or ‘construal’ in Cognitive Grammar's terminology) taking an intersubjective perspective. In the present treatise, the Cognitive Grammar descriptive approach is included in the discussion of the manner by which pre-linguistic intersubjectivity manifests itself in non-objective meaning.
2. THE INTERSUBJECTIVE BASIS OF NON-OBJECTIVE MEANING
The phenomenological notions of intentionality and intersubjectivity constitute the main concepts according to which non-objective meaning is explained in this paper. This section thus first presents the motivation for selecting these particular concepts (Subsection 2.1) and then explains these concepts in detail (Section 2.2). It is in place, however, to give a preliminary definition for intersubjectivity, for the content of this notion helps to underline the necessary interdependence between the social grounding of a linguistic expression and its non-objective semantic character.
Intersubjectivity is understood here essentially as a multi-level concept that subsumes the notions of intersubjectivity as (i) capability, (ii) joint activity, and (iii) an ontological category (Zlatev Reference Zlatev2008b, Möttönen Reference Meltzoff and Keith Moore2016):
-
(i) Intersubjectivity is the sharing of experiential and representative states either in an immediate embodied or in a symbolically mediated manner.
-
(ii) Intersubjectivity is not a unitary concept, but a functionally defined category for processes, structures and properties of human action which characterize joint intentional action and knowledge.
-
(iii) Categorizing a process, structure or property of human action or knowledge as intersubjective makes it irreducible to individual cognition or activity.
The main implication of this definition is that intersubjectivity involves detection and mediation of intentional states, i.e. acts and orientations directed toward external objects and states of affairs. Thus defined, intersubjectivity in fact comes with a structure provided by non-objective relations: orientations to extramental entities and states of affairs as intended from various subjective perspectives. Non-objective linguistic meaning, accordingly, is defined here as intentionality conventionalized into linguistic meaning. In what follows, I will argue that this property is inevitable for linguistic meaning to occur at all. The reason for the constitutive role of intentionality is best prefaced by addressing the social ontology of language and linguistic meaning, an extensive analysis of which was formulated by Itkonen (Reference Itkonen1978, Reference Itkonen, Haukioja, Helasvuo and Miestamo1997).
2.1 The social ontology of linguistic meaning
Itkonen's (e.g. Reference Itkonen1978, Reference Itkonen, Haukioja, Helasvuo and Miestamo1997, Reference Itkonen2008b) thorough analysis of the social grounding of language shows that linguistic meanings are constituted by successful practices of communication, so that they come to exist as conventional and social entities. Itkonen (e.g. Reference Itkonen, Haukioja, Helasvuo and Miestamo1997) argues that meanings exist as objects of common knowledge,Footnote 2 which then by definition, are shared by those who have become accustomed to that convention.
Common knowledge, therefore, has necessarily three levels (Itkonen Reference Itkonen, Haukioja, Helasvuo and Miestamo1997:55):
-
A knows x,
-
A knows that B knows x,
-
A knows that B knows that A knows x.
According to Itkonen (Reference Itkonen, Haukioja, Helasvuo and Miestamo1997), this three-level organization is directly applicable to our knowledge about language. Thus, ‘x’ above may be replaced with any valid expression in any language. The status of my example above as a correct English sentence depends on its status as an object of three-level knowledge. It is required that A (a speaker of English) knows that B knows that A knows that The young man that I once knew is now nothing more than a petty linguist is indeed a correct sentence (in English).
It is clear that such knowledge does not usually occur amidst linguistic interaction as an object of conscious reflection, but rather that this knowledge is presupposed by virtually any linguistic act. For an utterance to be uttered, it has to be known by A (Level 1). For the particular utterance to be a sensible choice for expressing A's communicative intention, that utterance has to be known by B (Level 2). However, for the act of uttering to be sensible in the first place, it needs to be understood as an expression of A's basic communicative intention – thus, A's knowledge (or expectation) about the explicitness of his or her own communicative intention is presumed by his or her act in the first place (Level 3).
Itkonen (Reference Itkonen, Haukioja, Helasvuo and Miestamo1997) shows that the basic three-level structure of common knowledge is also applicable to linguistic meaning. Indeed, common knowledge that would pertain only to linguistic vehicles (spoken or signed artifacts) stripped away from semantic content would be utterly useless from a communicative point of view. Language is acquired either as expressions or schematizations for multiple expressions, and those expressions involve both a vehicle and content (meaning).
If linguistic communication indeed does build on certain conventions or norms that characterize correct language use, it becomes inevitable that it also builds on conventional or normative meanings that are commonly known by the members of that speech community. Moreover, pragmatic (i.e. context dependent) meanings typically require pre-existing linguistic conventions in order to be graspable and understood. Obviously, the fact that these meanings are graspable pertains to the fact that they are available to the present interlocutors as facets of expressive acts, which, in turn makes them describable in terms of the three-levels of common knowledge as well.
This social character of linguistic meaning may seem self-evident at first glance, but it is often overlooked (e.g. Langacker Reference Langacker2008:27–30), neglected (Evans Reference Evans2009), or even denied (Johnson Reference Johnson1987:xxx–xxxi) by many semanticists (for further discussion, see Itkonen Reference Itkonen, Haukioja, Helasvuo and Miestamo1997; Möttönen Reference Meltzoff and Keith Moore2016:Chapter 3). I argue that this misappreciation of social is not only unfortunate from a theoretical point of view, but also detrimental to the very process of semantic analysis. In Cognitive Linguistics, the ontological commitment that meanings are actually intramental entities has led to the description of non-objective meaning as a correlate of cognitive activity or as a property of an internalized representation (see e.g. Langacker Reference Langacker2008:30, 79–80). In and of itself, the explanation of a semantic phenomenon with regard to a cognitive one is not at all infelicitous: human cognitive and bodily disposition certainly do have an effect on how expressions develop meaning for us. For instance, the semantic content of an expression is necessarily delimited by our cognitive processing capacity – albeit, it is necessary to keep in mind that this delimitation comes into being exactly via intersubjectively grounded linguistic acts.
In contrast, problems arise when meaning is reduced to mere cognitive activity, so that the internalization or individual conception is assumed to constitute the meaning as a whole. As any cognitive entity per se is not directly graspable or fit to constitute an explicit intentional relationship between a subject and an object of expression, such an entity can be linguistically relevant only inasmuch as it is shown to manifest through the intersubjectively available contents of an expression. In other words, linguistic meaning is representative of its designatum, and the particular way of representing a designatum is necessarily determined by the ability of the human mind to conceptualize it. Linguistic representation, however, is not a cognitive representation, but rather part of an intentional relationship that as a whole constitutes that meaning. In other words, the meaning of a linguist is not the mental representation of a linguist, but rather a particular, intersubjectively shared way of intending a linguist.
The social ontology that Itkonen provides for linguistic meaning manifests a usage-based conception of language. Expressions and grammatical rules are not only learned via use; they also exist as instances and constants between those instances (e.g. Itkonen Reference Itkonen2008b:285–286). It is thus paramount that meaning is not confused with the knowledge about meaning that the member of a speech community possesses. Instead, meaning is an object of such knowledge (e.g. Itkonen Reference Itkonen, Haukioja, Helasvuo and Miestamo1997:63).
While such a usage-based position is accepted here as a valid ontological delineation, I interpret its implications somewhat differently from Itkonen. Namely, Itkonen is willing to interpret non-objective, representational facets of meaning as entities of private experience and thus as strictly distinct from linguistic meaning as a social entity (e.g. Itkonen Reference Itkonen2008a:23). In contrast, I argue that non-objective meaning is a socially valid and integral part of linguistic acts. The motivation for taking this point of view comes from a phenomenological analysis of linguistic meaning.
2.2 The phenomenological basis of semantic analysis: Intentionality and intersubjectivity
As is argued persuasively by Itkonen (Reference Itkonen and Perry1980, Reference Itkonen2008a) and Zlatev (Reference Zlatev, Ziemke, Zlatev and Frank2007b, Reference Zlatev2008a), linguistic meaning qua social entity needs to be consciously accessible in order to be meaningful for ordinary language users or analyzable by linguists. If intersubjectively valid use constitutes linguistic meaning, meaning must reside in concrete acts of use. These acts, in turn, exhibit various intentional relationships. From the perspective of the utterer, there is the intended meaning and, typically, the experience of a foreign subject, who is capable of sharing the original intention from his or her own perspective.
Phenomenology, in brief, is a methodology developed for revealing the aspects, processes, and structures of consciousness that make coherent and thus meaningful experiences possible. I will concentrate on the two phenomenological notions developed especially by Husserl (Reference Husserl2001a, b; Zahavi Reference Zahavi2001a, b, Reference Zahavi and Thompson2003), which I consider necessary for a precise theoretical depiction of non-objective linguistic meaning, namely, intentionality and intersubjectivity.
As mentioned already in the introduction, intentionality is the inherent object-directedness of consciousness. Whether we consciously concentrate on an object or do not, our experience is almost invariably about something that stands as separate from our consciousness per se. In phenomenology, intersubjectivity refers to the capacity to experience foreign intentional subjects and identify with them.
The phenomenological analysis of intentionality assumes, as its primary challenge, to unveil how various types of intentional acts (perceptions, interactions, expressions, etc.) come to hold meaning for us as cognizant subjects. A particularly relevant part of these acts, from a linguistic point of view, is the layered character of intentionality and its objects. The meaning that extramental objects and our everyday interactions with them have for us is provided by the synthesis of (an indefinite amount of) previous interactions (Drummond Reference Drummond2012:123; Gallagher & Hutto Reference Gallagher and Hutto2008:26–27). For example, a mug does not appear to us as ‘a mug’ because of its inherent properties, but rather because of its properties that have emerged from our previous experiences of seeing, feeling, handling, tilting, and drinking from various mugs.
However, even the experienced entirety of a mug as a physical artifact requires a synthesis of previous sensuous experiences. Husserl demonstrates the necessity of such synthesis in his analysis of what he calls the immanent content of experience (Husserl Reference Husserl2001a:99; Zahavi Reference Zahavi and Thompson2003:22). The immanent content of experience refers to the stuff of sensuous experience prior to its organization into meaningful objects, events, and states of affairs. For instance, nothing in the visual characteristics of an object alone, such as hue, luminosity, or boundaries vis-à-vis the background, is enough to constitute an object as an autonomous entirety. Rather, it is the active (also perceptually active) interaction with the entity that leads to the constitution of an object in our consciousness, as a synthesis of our interactions with the entity and the appearances it may have had in them.
Thus, our conscious experience is typically governed by meaningful entities, whereas the raw immanent content of our experience can be grasped only with a considerable amount of additional effort. My consciousness is constantly attached to an objectivity that involves the implicit integration of multiple perspectives. This integrative nature of experience implies the integration of perspectives between multiple subjects. That is, when I visually intend to see a mug as a genuine object, I also intend it as being the same object for another, co-present subject; despite the fact that this other individual intends the mug through a different perspectival appearance. A significant portion of intentional acts and their objects thus has an intersubjective basis, even when experienced privately.
The existence of other subjects is presupposed in our everyday experience of the reality of objects, but also by the very existence of the objects as something that we experience simultaneously as a whole and as perspectival (Zahavi Reference Woelert1997:312). Intersubjectivity, however, is not constituted by intentionality itself. Rather, the human capacity to intersubjectivity is a multifaceted property of experience.
Significant findings about infant intersubjectivity have been made by developmental psychologists who were inspired by phenomenology. While newborns and small children cannot give verbal accounts of their experiences, there is mounting evidence that children actually possess substantial pre-linguistic intersubjectivity skills (see e.g. Astington Reference Astington2006; Gallagher & Hutto Reference Gallagher and Hutto2008; de Bruin & de Haan Reference De Bruin and de Haan2012; Meltzoff & Moore Reference Lewis1977, Reference Meltzoff and Moore1994, 1997; Stern Reference Sinha, Jansen and Redeker1971, Reference Stern1977, Reference Stern1985; Trevarthen Reference Tomasello1979, Reference Trevarthen and Bullowa1980; Trevarthen & Aitken Reference Trevarthen, Olson, Bruner and Miller2001). They are likely to possess the capacity to tune to caretaker's actions and sense a caretaker's intentions without making action-based inferences about her psychological states (e.g. Trevarthen & Aitken Reference Trevarthen, Olson, Bruner and Miller2001:7). Instead, the children experience foreign intentionality as an explicit, constitutive property of that other individual's actions (Gallagher & Hutto Reference Gallagher and Hutto2008:22).
Zlatev (e.g. Reference Zlatev2008b) presents a layered model of intersubjectivity, according to which linguistic representational capacity develops based on more rudimentary and direct forms of intersubjectivity and related mimetic skills. Zlatev's model includes different developmental stages that roughly correspond to increasingly mediated forms of intersubjectivity (Zlatev Reference Zlatev2008b:218–221). These forms, in turn, are defined by the extent to which one (an infant) is able to coordinate his or her own intentions with those of the Other (e.g. a caretaker). For instance, small children from nine months onward show the capability of a type of triadic intersubjective coordination called joint attention, as to simultaneously sense and coordinate the intentionality of the Other toward an external object – ‘I see that you see X’ – and to recognize the mutuality of that intention: ‘I see that you see that I see X’ (Zlatev Reference Zlatev2008b:239–240).
It is not accidental that joint attention bears the same structure as common knowledge, which was discussed above. This kind of triadic perceptual intersubjectivity is necessary for linguistic intersubjectivity to occur (Zlatev Reference Zlatev2008b) because the necessary conditions for triadic intersubjectivity, such as joint attention, subsume not only the capacity to recognize the Other's intention, but also the reciprocal understanding that this intention is explicit in one's actions.
Whereas there is clearly a strong innate basis for recognizing foreign intentionality, it is also clear that situated bodily interactions with the caretaker promote the ability to use one's body (e.g. the direction of gaze) to mediate intentions, which in turn grounds both the understanding and the production of linguistic signs. Once the child's capacity to triadic intersubjectivity is in place, it is ‘only’ required by signitive action that the mutual recognition of intentionality is channeled via a (mutually recognized) sign toward the intended object. That is, such signitive action is extended from strongly situated bodily expressions to more and more non-situated or conventional expressions, such as actual linguistic verbalizations.Footnote 3
Thus, there are strong empirical and theoretical grounds that suggest that language is an extension of pre-linguistic intersubjectivity, which cannot develop in the absence of interaction with other subjects. What is significant about this view is that the pre-linguistic intersubjective grounding of language actually suggests the intersubjectivity of language itself. Meanings that are constituted in an intersubjective setting based on understanding and coordinating the intentions of Others involve, by default, their sharedness between multiple subjects. Quite remarkably, experientially learned meanings are thus readily compatible with Itkonen's formulation of social ontology as a three-level common knowledge. The next task is to demonstrate how these theoretical premises can be integrated into a precise description of non-objective meaning.
3. THE PHENOMENOLOGICAL APPROACH TO NON-OBJECTIVE MEANING
The compatibility of the phenomenological notion of intersubjectivity with Itkonen's (Reference Itkonen, Haukioja, Helasvuo and Miestamo1997) social ontology is the basic prerequisite for explaining non-objective meaning theoretically. Most importantly, the capability of coordinated intentionality found in pre-linguistic children suggests that linguistic meaning thus constituted may assume the basic structure of intentionality. That is, if we are pre-linguistically able to detect and coordinate intentions, there is no a priori reason why the manner in which one intends an entity could not conventionalize itself into the meaning of an expression that refers to the same entity in question. In fact, I argue that linguistic meaning is necessarily characterized by human intentionality and the facets of communication that derive from that intentionality. The full explication of this position requires that the phenomenological concept of meaning-intending action is introduced and applied to the usage-based conception of language.
3.1 Non-objective meaning in meaning-intending acts
Above, intentionality is presented as the object-directedness of consciousness that constitutes the object as the object of consciousness. For example, the heaping pile of unread books that I observe when glancing across a room is constituted as an object by my perceptual intention. This constitutive function of intentionality can hold for linguistic meaning as well, namely, when I actually speak of ‘that heaping pile of unread books’. However, the type of intentionality that manifests itself in linguistic meaning shows characteristics that have not yet been fully explicated.
Husserl (Reference Husserl2001a, b) offers a systematic description of a wide array of different intentional acts, but those that relate to symbolic mediation are analyzed using the rubric of meaning-intending. In brief, acts of meaning-intending involve consciousness, which transforms a physical sign into an expression, i.e. they involve a relationship between the sign and its designatum. What is noteworthy, however, is that the same expression-status is conferred by this relationality itself. Even an iconic sign, e.g. a bodily imitation or a picture, is understood as an expression only inasmuch as it is grasped as a correlate of foreign intentionality, that is, the intention of the Other, who intends the vehicle to be associated with its designatum.Footnote 4 Consequently, meaning-intending acts also contain content that represents the intersubjectively valid and explicit facet of the intention (see Banchetti-Robino Reference Banchetti-Robino1997:311–312).
It is thus the content of a meaning-intending act that corresponds to its meaning. That is, when I say that heaping pile of unread books, my expression is about a particular pile of books as an extra-linguistic designatum; however, the meaning of the expression is ‘the pile of books’ as intended in a particular manner (Banchetti-Robino Reference Banchetti-Robino1997:311). That is, the meaning cannot be the intentional object itself, for in that case there were now difference between linguistic meaning and a meaningful experience as such.
It is clear at once that meaning thus defined is inherently non-objective, albeit it is also dependent on a genuine extralinguistic object: it involves the intentional relationship between the subject and the object of the expression. The content of meaning-intending acts is not subjective either. Genuine meaning-intending acts are grounded in actual linguistic practices that presuppose their own validity for multiple subjects. In order to make sense, I must succumb to a conventional manner of intending. The notion of a meaning-intending act is, therefore, quite compatible with Itkonen's formulation of meaning as an object of social knowledge.
From a phenomenological point of view, then, the meaning of an expression can be defined as the content of a meaning-intending act that corresponds to a particular expression. The problem here is that the notion of ‘content’ remains somewhat vague. Meaning/content is what is interpreted as an intention, but the conventional/normative character and relative abstraction of linguistic meanings require that this content is also determined by convention of use. However, I argue that the content and use of the expression also pertain to distinct categories. That is, a linguistic meaning-intending act involves a genuine linguistic representation, wherein the vehicle stands for the designatum by selecting and organizing the content that pertains to it. In fact, it is this organizing, non-objective facet that necessitates the postulation of meaning as a category that is inherent to, but also discernible from, actual use.
First of all, originating in intersubjectively valid linguistic intentions, the content of a meaning-intending act is characterized by multiple linguistic representations of the designatum or designata, which are interpretable as representatives of the same category. Thus, the actual use of an expression determines its meaning, but that meaning resides in an explicit intentional relationship between an intentional subject and an intentional object. This relation represents information about the object by portraying it in a particular manner. Second, co-referential expressions seem to bear different non-objective contents. That kettlebell or that useless lump of iron may be used co-referentially but these expressions nevertheless differ in meaning. The difference is obviously a result of different ways of using these expressions, but it is also explicit in the attributes that these expressions assign to their shared designatum. Third, using these expressions differently in a motivated manner requires there be a criterion according to which either of these expressions is selected. The most self-evident candidate for that criterion is the manner in which the expressions represent the properties of designatum and, consequently, the relationship between the utterer and the designatum.
Thus, it seems reasonable to argue that there is representational non-objective meaning inherent in meaning-intending acts. While it is the intention itself that constitutes the representational relationship between the vehicle and the designatum, the non-objectivity of the meaning requires that this relationship be contentful. Moreover, this content is not reducible simply to a mental representation in any simulative or conceptual sense – although such representation may co-occur along with a meaning-intending act. Rather, the content is derivative of multiple normatively correct meaning-intending acts that have already been successfully interpreted.
Ultimately, the non-objective representational meaning can be presented as a logically necessary property of linguistic function. Intersubjectively valid intentions ground linguistic expressions. An intention, in turn, always corresponds to a relationship between a subject and an object, so it necessarily involves non-objective characteristics.Footnote 5 The non-objective aspects of meaning are thus attributable to the content of the meaning-intending act.
3.2 A usage-based approach to language and its implications
At the most rudimentary level, usage-based grammar is a system of linguistic units that are constituted by, and acquired in and through actual communication. In so being, a usage-based theoretical model of grammar typically involves the interpretation of different grammatical phenomena, including syntactic constructions, as being inherently meaningful. Cognitive Grammar (Langacker Reference Johnson1987, Reference Langacker2008) is a paramount example of this particular approach, as are the different varieties of construction grammar (e.g. Goldberg Reference Goldberg1995, Croft Reference Croft2001). In general, usage-based theories of grammar place significant emphasis on internalization, or the cognitive ‘entrenchment’ (Langacker Reference Johnson1987:59) of linguistic units. A particular usage-based model can be assessed relative to the role attributed to internalization in the overall conception of language.
For Cognitive Grammar, it is exactly the linguistic units, viewed as being internalized, that become the primary subject of linguistic inquiry. Cognitive Grammar defines linguistic meaning as conceptualization, that is, as cognitive events that correspond to the activation of conceptual content (Langacker Reference Langacker2008:30). On the other hand, Cognitive Grammar accepts the fact that conceptual content is acquired by exposure to how linguistic units are used in various contexts (Langacker Reference Langacker2008: 457–458). Thus defined, internalized units correspond to generalizations (i.e. schemas, e.g. Langacker Reference Langacker2008:215) that capture the similarities evident in the natural linguistic ‘data’ to which the language speaker is exposed.
These two perspectives may appear to fit together quite smoothly, as overt constants of language use result in covert patterns of linguistic processing; the latter may then be recruited for the interpretation of new linguistic occurrences. There is an inherent tension between overt linguistic use and covert cognitive entities, however. First of all, intersubjective validity is a necessary characteristic of a linguistic sign that cannot be grounded in private cognitive phenomena. Second, linguistic signs are used to convey meaning about extralinguistic, extramental reality, a function that again cannot be grounded in any private cognitive phenomena. Thus, language cannot be reduced to its internalization.
The tension thus lies in the fact that internalized linguistic knowledge is both necessary and insufficient as the basis of linguistic meaning (and language). However, it is resolved by accepting internalized linguistic knowledge as a part, but only a part, of that linguistic meaning that is defined by use and not vice versa. This understanding, in turn, brings us to the issue of compatibility between the phenomenological analysis of meaning and a usage-based account of grammar.
What is overlooked by many usage-based models is that language acquisition presupposes the capacity for interpreting linguistic intentions. An expression is not interpreted as a correlation between sound (or writing) and a non-linguistic entity, but rather as an expressive act committed by a genuine subject. Thus, what is acquired in exposure to occurrences of a specific expression cannot be reduced to some representational content per se; rather, the internalized knowledge about learned language assumes as its object the manner in which any expression is used. Any coherent usage-based conception of meaning thus needs to involve the detection and interpretation of the intentionality present in an expressive action. This necessity in turn makes it feasible to explain non-objective meaning in terms of conventionalized intentionality; this explanatory relationship is the gist of the present approach.
The challenge of a usage-based approach is to determine the relationship between use and conventionalization, on the one hand, and between conventionalization and internalization, on the other. It is clear that a certain amount of internalization is necessary for meaningful linguistic communication (see also Etelämäki & Visapää Reference Etelämäki and Visapää2014). In particular, the semantic complexity of the most mundane discussions suggests that linguistic meanings rely to a substantial degree on complex and conventionalized patterns of conceptualization (see Section 3.1 above). On the other hand, by being conventional, these patterns are explicit in the expressions that manifest them. It is, therefore, reasonable, from a theoretical point of view, that the role of internalization should be minimized according to the minimum requirements of a given theory.
I apply this principle here in the following manner. I suggested above that the inclusion of non-objective meaning should be considered as a part of the criteria according to which an expression is selected for its particular use in a particular context. For a usage-based conception of language, this principle suggests that non-objective meaning should also be considered part of the more conventionalized and internalized semiotic repository of language. Nevertheless, this possibility does not in any way challenge the basic tenet of Itkonen's (Reference Itkonen, Haukioja, Helasvuo and Miestamo1997) version of the usage-based approach to linguistic meaning, as in principle, any characteristic of cognition may be linguistically relevant only inasmuch as it is manifested in genuine linguistic acts.
The combination of intentionality-based and usage-based approaches suggests that non-objective meaning is a genuine representational facet of meaning. In fact, it can be argued that using an expression in a particular manner presupposes a particular non-objective representational content exactly as part of its use. That is, the criterion according to which a particular expression is selected for any use must pertain to the manner in which that expression conveys meaning between multiple subjects. Given the necessary requisite that linguistic meaning is socially valid, non-objective meaning must also exist as a representation of intentionality as conventionalized into actual expression.
In this formulation, the representational non-objective meaning is considered an inherent part of use; that is, part of concrete, linguistic meaning-intending acts. This ontological delineation is best illustrated by focusing on a particular type of non-objectivity, as analyzed in Langacker's Cognitive Grammar. At the same time, this analysis brings our focus back to linguistic intersubjectivity.
4. INTERSUBJECTIVITY AND CONSTRUAL
Cognitive Grammar scrutinizes non-objective meaning as discernible of various dimensions of construal (Langacker Reference Langacker2008:55–93). In the context of Cognitive Grammar, construal can be interpreted as those properties of conceptualization that stem from the activation of conceptual content for a single representative speaker. The dimensions of construal are then used to discern co-referential expressions that do differ in meaning. Thus, alternate construals for the same referent, an object or a state of affairs, pertain to the different ways of conceptualizing the referent in question.
The present approach to non-objective meaning accepts the descriptive validity of the dimensions of construal, but defines construal as a facet of the content of a meaning-intending act. This redefinition, I argue, is non-trivial for multiple reasons. First of all, as a facet of intentional content, non-objective meaning – or construal – is dependent on a non-construed and ultimately non-linguistic intentional object. This aspect is necessary for non-objective meaning to exist as an intersubjectively valid phenomenon in the first place. Second, non-objective meaning could not exist in its communicative significance if it were not detected against the backdrop of the signified entity or state of affairs as the objective (i.e. non-construed) whole. The very idea of construal as a means of expressing ‘the same situation’ (Langacker Reference Langacker2008:43) in alternate ways presupposes that ‘the same situation’ is an integration of the realized and potential construal alternatives. This aspect is what makes construal non-trivial, both as a manifestation of non-objectivity and a mediation between the different perspectives from one construal to another.
What follows from these remarks is that construal can be attributed to two mutually complementary functions. On the one hand, construal pertains to the non-objectivity of meaning. On the other hand, the linguistic act of constru-ing involves selecting an expression according to its non-objective meaning and according to the discursive context. Thus, construal has a reflexive character, which we can illustrate by considering just one specific dimension of construal.
The most one-dimensional of the dimensions that Cognitive Grammar posits is the dimension of specificity. This dimension can be exemplified by identified natural taxonomies formed by lexemes, i.e. dog → terrier → Cairn terrier (the arrows indicate the increase in specificity). These expressions are applicable to the same entity, given that this entity fulfills the characteristics of the most specific expression, i.e. Cairn terrier. This last expression is the most specific of the three in that it defines the properties of its designatum to a greater extent than do the other two.
Cognitive Grammar's explanation for the increase in specificity from an expression such as dog to Cairn terrier involves the recourse to a conceptualization, so that a more specific expression corresponds to a conceptualization with more defined inherent characteristics (Langacker Reference Langacker2008:56–57). The non-cognitivist approach that is preferred here, however, requires that the relative specificity of an expression is related to acts of meaning-intending. These three above-mentioned expressions may be considered alternate construals for the same designatum, indeed correlates of different meaning-intending acts that share the same intentional object, but differ relative to their specific content. The explicit representative function of these expressions pertain to the acts that relate their content to the designatum. Specificity can therefore be described as that level of detail according to which the meaning-intending act predefines the characteristics of the designatum.
The specificity of an expression is a result of the manner in which that expression is applied to various designata, which makes specificity an inevitably intersubjective dimension. As a usage-based feature, specificity (or given value thereof) must be a commonly known property of the expression in question. For instance, it is a commonly known property of the lexical concept ‘dog’ that it subsumes breeds that subsume breeds and so on. At the same time, a detailed taxonomy of canines is not likely familiar to the majority of any speech community. However, the meaning ‘Cairn terrier’ exists due to a conceptual/linguistic sub-community (e.g. dog owners and Cairn enthusiasts) whose communicative purposes dictate the referential scope of that particular notion.
The preceding discussion on intersubjectivity, however, has further implications for the dimension of specificity. The fact that the expressions dog, terrier and Cairn terrier may be used interchangeably for the same designatum implies the possibility that each use is an instance of motivated selection. In other words, there are reasons, albeit typically pre-reflective, for uttering dog instead of Cairn terrier or vice versa. As such a selection aims at maximally effective (i.e. intersubjectively graspable) communicative act, the main criterion for the selection is to prompt an interpretive meaning-intending act that corresponds to the original intention of the choice. In other words, the intersubjective setting of a communication constitutes the chief determinant of construal.
The dimension of specificity shows a particular type of context-reactivity. The selection is dictated by the extent to which the interlocutors share a common knowledge base. This shared knowledge-base is interpreted by the interlocutors as an overlap of their subjective knowledge-basis. The construal in terms of specificity then pertains to the extension of shared knowledge on the basis of the preceding discourse and the overall context of the utterance.Footnote 6 The dimension of specificity can thus be given a definition that corresponds directly to the intentionality-based model developed here. As a property of a meaning-intending act that is based on linguistic convention, the specificity of an expression pertains to construal as a means to establish intersubjectively shared access to the intended designatum.
Construal in terms of specificity presupposes three distinct types or levels of intersubjectivity. First, it involves the conventional, socially defined specificity of an expression. Without this conventionality, there would be no criterion according to which the extra-linguistic designatum could be construed relative to specificity in the first place. This implies a second type of intersubjectivity relevant to specificity and a construal at large. The act of construing involves the alignment of different perspectives that correspond, in Cognitive Grammar terminology, to distinct conceptualizers (Langacker Reference Langacker2008: Section 12.3.2; see also Verhagen Reference Verhagen2007). A construal that corresponds to a given expression may, therefore, have conventionalized effects that adjust the relationship between the utterer and the recipient. In the case of specificity, the informational value of a conventional linguistic expression is, by default, symmetrical between the members of the speech community. The well-demonstrated (e.g. Verhagen Reference Trevarthen and Aitken.2005) existence of conceptualizer-specific effects, however, proves that a symmetrical construal is a motivated choice.
The symmetry of construal in terms of specificity leads one to the third type of intersubjectivity that is relevant for non-objective meaning. However automatic and pre-reflective it may be, construal is realized in the actual use of linguistic expressions, relative to which it can be defined as a motivated selection of expression/s in an intersubjective context. Specificity, for instance, is only a functional facet of an expression inasmuch as it corresponds to a relevant selective criterion in certain contexts. Furthermore, any context itself is inevitably and non-trivially intersubjective, so it only makes sense to choose an expression with regard to specificity inasmuch as that choice may direct the recipient's interpretation of the intended outcome.
Despite the relative abstraction of their definitions, these three types of intersubjectivity can be considered as necessary a priori conditions to have a coherent notion of non-objective meaning, regardless of the particular construal dimension (see Möttönen Reference Meltzoff and Keith Moore2016). Now that these types have been explicated, however, they can be applied to consummate the present account of non-objectivity as a facet of actual meaning-intending.
First and foremost, the explication of the types of intersubjectivity relevant to non-objective meaning underscores the fact that non-objective meaning involves a particular type of semantic effect: non-objective meaning is intersubjective in that it is explicit in any given or offered expression. Inasmuch as the use of the expression in question is motivated by some pragmatic factor, that expression's non-objectivity stands as a criterion for its selection. This possibility of pragmatic selection implies, in turn, the existence of a context-independent, representational, non-objective meaning that can serve as the criterion.
Finally, this context-independent, non-objective meaning does in no way depend on a particular type of cognitive representation. In contrast, it merely presupposes a capability of representing an extra-linguistic entity correctly by using a particular expression; it is a capability that in and of itself has a representational function. Non-objective meaning can thus be associated with the capability to intend linguistically in a particular manner, which in and of itself can involve internalization. Internalization by itself does not constitute context-independent meaning; rather, it makes it possible for context-independent meaning to be actualized in each particular instantiation of such meaning.
5. DISCUSSION
Above, I have considered the necessary pre-linguistic intersubjective basis for non-objective meaning. With my brief excursion into the phenomenological concepts of intentionality and intersubjectivity, I aimed to demonstrate that non-objective linguistic meaning can be explained and described as one of the properties of linguistic meaning-intending acts. Such an approach is not only concomitant with the social conception of meaning à la Itkonen, but also can be derived from a usage-based account of language. As illustrated in Sections 3.2 and 4.0, the approach developed here diverges from the standard cognitive linguistic explanation of non-objective meaning via a postulation of intramental structures. Rather, it underscores the dependence of non-objective meaning on extramental objects, relative to which different construals and perspectives can be well aligned. Consequently, non-objective meaning is only a sensible feature of language only if that meaning pertains to an explicit, intersubjectively validated manner of intending.
It is this property of sharedness that makes construal relevant for the analysis of discourse. That is, the manner in which a meaning-intending act construes an entity or a state of affairs non-objectively, can be defined as a criterion of selection – or a source of semantic motivation. Construal, thus defined, does not exhaust semantics, let alone pragmatics (when considered from the perspective of representational content). It can nevertheless be argued that, for any usage-based semantic theory, non-objectivity is the central facet of representational meaning. Since non-objectivity does not stand by itself, but rather motivates pragmatic selection according to various dimensions, it is clearly a promising starting point for a new synthesis between the interactional and cognitive approaches to language. Despite the theoretical emphasis of the present treatise, however, I do hope it has succeeded in demonstrating the empirical relevance of this task.
ACKNOWLEDGEMENTS
I wish to thank both the editors of the present special issue of NJL and the anonymous referees for the valuable feedback that helped me to clarify the main arguments and the overall structure of this paper.
