Reading Time: 24 minutes

The meaning-sharing network

John Haglund

Department of Philosophy, Lund University

Johan Blomberg

Centre for Languages and Literature, Lund University


Abstract. We advocate an analysis of meaning that departs from the pragmatic slogan that “meaning is use”. However, in order to avoid common missteps, this claim is in dire need of qualification. We argue that linguistic meaning does not originate from language use as such; therefore we cannot base a theory of meaning only on use. It is important not to neglect the fact that language is ultimately reliant on non-linguistic factors. This might seem to oppose the aforementioned slogan, but it will be made clear how this opposition is chimerical. We propose that meaning traces back to the relation between subjectivity and intersubjectivity, which is at the heart of the matter by inducing strong interdependency between intention and interpretation. But to base a full-fledged analysis of meaning in communicative dyads alone is also insufficient. What needs to be further acknowledged is that meaning sharing becomes regulated by the interactions of a community. This we call consensus and it is at play in framing all communicative acts. Hence arises a triadic structure – we call this the meaning-sharing network. The main motivation behind this model is to capture that meaning is not “in the head”, nor is it autonomous of the individual members that constitute it.

Keywords: meaning, pragmatics, philosophy of language, communication, intersubjectivity

Tähenduse jagamise võrgustik

Artiklis esitatakse tähenduse analüüs, mis tõukub pragmaatilisest loosungist „Tähendus on kasutus”. Selleks, et vältida levinud vigu, vajab antud väide kindlat piiritlemist. Me väidame, et lingvistiline tähendus ei tulene keelekasutusest kui sellisest, mistõttu ei saa rajada tähendusteooriat ainult kasutusele. Oluline on mitte jätta tähelepanuta asjaolu, et keel põhineb mittelingvistilistel faktoritel. Võib tunduda, et see väide on eelmainitud loosungiga vastuolus, kuid artikkel selgitabki antud vastuolu kimäärset olemust. Tähendus põhineb subjektiivsuse ja intersubjektiivsuse vahelistel suhetel, mis omakorda tingivad intentsiooni ja interpretatsiooni vahelised tugevad seosed. Samas on üksnes kommunikatiivsetele diaadidele rajatud tähendusanalüüs ebapiisav. Oluline on, et tähenduse jagamist reguleerivad kogukonna interaktsioonid, mida me nimetame konsensuseks. Need mängivad olulist rolli kõikide kommunikatsiooniaktide kujunemisel. Nii moodustub triaadiline struktuur, mida nimetame tähenduse jagamise võrgustikuks. Antud mudeli loomist on ajendanud tõdemus, et tähendus ei asu „peas” ega ka lahus indiviididest, kes selle tähenduse moodustavad.

Märksõnad: tähendus, pragmaatika, keelefilosoofia, kommunikatsioon, intersubjektiivsus


The assertion that language has an intrinsic power to refer to and mean something about the external world; that there is a close and mind-independent correspondence between language and reality, is a clear tendency in philosophy of language. All the way from Meinong via Frege and Russell to the present day this tends to lead to reification of some certain aspect that provides propositions with sense. That is, in one way or another there just has to be such a thing as the meaning. As observed by Priest, history provides ample proof that the search for such a correlate with few or no exceptions ends in contradiction. [1]

It is natural for a theory of language to have implications about what can and cannot be expressed. […] Modern theories of language always seem to render something – usually themselves – beyond the limit of expression. (Priest 2002: 179)

We present an analysis of meaning inspired by the slogan “the meaning is the use”, often attributed to the later Wittgenstein. This claim is however far from clear, and Wittgenstein’s own formulation is left open to interpretation.

For a large class of cases – though not for all – in which we employ the word “meaning” it can be defined thus: the meaning of a word is its use in the language. (Wittgenstein 1953: §43)

If meaning is the use, then this reasoning might at the face of it lend itself to conclusions close to behaviorism or scepticism. There is, on the other hand, another traversable route available for the sensible interpreter that wishes to protect the pragmatic approach from such precipitate conclusions: we present a version of this pragmatic slogan that situates linguistic meaning in a wider context of how subjects generate, and share, meaning. As soon as we step outside the confines of language, we see that there are certain preconditions to linguistic meaning. What should also become clear is that linguistic meaning is only part of the broader category of what we call signified meaning. Signified meaning includes, besides language, every other form of sign use and in line with this we will introduce the term sign game, originating from “language game” but extended to apply to all types of sign use. Do note that throughout this paper we will understand the sign as something that stands for something for someone, the point being that there is no meaning but meaning for a subject. Another way to phrase this is to follow Sonesson’s definition of the sign as involving a differentiation between expression and content from the subject’s point of view. Thus, we do not endorse a theory of meaning as solely mediated by signs, in fact the only reason that signs have meaning at all is because they have been intended by someone and later becomes interpreted by someone.

Our suggestion is that signified meaning in turn ultimately depends on

(1) the organizing of experience of the lived world from the perspective of the subject, what we will dub intentional meaning.


2) the establishment of intersubjective meaning-sharing, both with and without the use of signs, where meaning-sharing in the absence of signs will be called joint experience, the intersubjective version of intentional meaning (which includes such phenomena as mutual gaze).

Since we live our lives collectively and since we all have broadly speaking the same constitution, this organization is carried out intersubjectively from similar vantage points. As we will argue, there are reasons as well to distinguish between intended meaning and interpreted meaning — both of equal importance. (1) and (2) above are necessary conditions for the possibility of signified meaning. As sign use through sign game playing develops into a larger system where meaning is constantly being transmitted, a matrix arises through which both intended meaning and interpreted meaning is filtered. This matrix of sign game regulating functions, which upholds a close link between agreement and rule-following we will call consensus; in this sense consensus is the general rule-following. Language in its full-fledged form is a prime example of such a larger structure where the sign system goes beyond the individual members of a linguistic community. Hence, a triadic structure between intention, interpretation and consensus takes form, similar to a first-person, second-person and third-person perspective, respectively. These make up the three dimensions of meaning. The members partaking in this flux of sign games should then be conceived of as both rule-following and rule-shaping, both as regulated and regulating [2]. Meaning is thus spread and distributed in an intersubjectively founded network of meaning-sharing, which cannot be reduced to its respective domains; rather, all significant levels must be taken into account. The aim of the present paper is to broadly outline the general framework of this meaning-sharing network.

Why meaning is, and is not, use

Language is a grand meaning-carrying sign system, and since we depart from philosophy of language and linguistics it is the natural exemplificandum. Nevertheless, it is important to keep in mind that language use is merely one case of sign use.

The slogan “the meaning is the use” states that linguistic meaning can neither be determined on the basis of reference to a class of designated objects in the external world nor can it ultimately be reliant on mental content, such as imagery or symbolic representations. It is therefore a philosophical fallacy to infer anything absolute about “evil” from the statement “Ivan the Terrible committed many evil deeds”. Rather, a given word, such as “evil”, attains meaning by its possible usages across different so-called language games.

From time to time, a tendency can be seen to take a pragmatic stance on meaning analysis to its logical extreme, the most probable explanation for such a misstep would be to forget that there always is a user behind the use. Serious ramifications will follow from such readings [3]. First off, to state that the transmission of linguistic messages is devoid of all but the here-and-now use is one step towards behaviorism or, perhaps even more rash, scepticism [4]. Consider language acquisition from the extreme philosophical behaviorist point of view: x learns that in a certain context, people use a certain phonetic string for some reason or another. Probed by the context, x starts to use this string in similar situations. The use is either accepted or rejected by the linguistic community. Dependent on which, x continues to use the noise in such contexts, or the use is revised. Such an analysis clearly leaves out some of the processes behind meaning formation and fixation; and, for the record, this behavioristic line was explicitly denied by Wittgenstein, albeit with a cryptic touch:

Are you not really a behaviorist in disguise? Aren’t you at bottom really saying that everything except human behavior is a fiction? – If I do speak of a fiction, then it is of a grammatical fiction. (Wittgenstein 1953: §307)

We wish to present another way of making sense of the later Wittgenstein, devoid of behaviorist pretensions. The proper way to read his later work, in our opinion, is that his focal concern was to reveal the limits of language as a tool for philosophical investigation. In other words, if language attains meaning through use, then analysis of language fashion will not unveil insights across this border. This is illustrated by the thought experiment of “the beetle in the box” (Wittgenstein 1953: §293). Imagine that everyone in a community possesses a box said to contain a beetle. The only evidence for there being beetles in the boxes is the fact that linguistic reports are given in support of this state-of-affairs, i.e. each and everyone can only peak into her own box. Language alone is incapable of verifying what any given box contains, the box might as well be empty; analyzing linguistic reports can neither serve as evidence for the existence of a particular beetle nor serve to confirm its non-existence. And yet, it is still valid from the viewpoint of the language game to speak of the beetle.

At the face of it, it might surely seem as if the behaviorist was right: there is no way to know anything apart from what people are actually saying and doing. But here, we disagree. It is one thing to note that a speaker is far from free to use language as she sees fit – language games are rule-bound in that not only sender but also receiver has to be aware of the intended/interpreted meaning. To us, “meaning is use” is an explication of what is going on in ordinary language use. But let us not forget that it is use for someone, in other words where there is usage there is necessarily a user. It is, therefore, impossible to detach use from the using subject and take the strict third-person view and the question about what motivates the rules of the language games is still open. Language is then like a game insofar as the rules for participating are commonly known (cf. Itkonen 2008) [5]. Meaning is then intimately tied to the activities of the users, individually and collectively. By using something in a certain way in certain contexts it is provided with meaning that is founded in that it is use for someone. We conclude that if meaning is use, and we believe that it is, then there must be something missing in the story. Consequently, the linguistic domain has to be founded in something else.

Beyond “meaning is use” — extra-linguistic preconditions for meaning

Our notion of meaning as use entails that signified meaning becomes fixated through the specific usages. Consider a linguistic sign game: the sign “elephant” attains its meaning in the first instance by being used to refer simply to elephants in the lived world. This analysis is adequate internal to the sign game context; however, something crucial is being omitted. As previously stated, to claim that the fixation of meaning in language is based on use does not entail an argument concerning how meaning as such can crop up in the first place [6]. To explain this we insist on acknowledging that there is meaning prior to the introduction of language, prior to any sign game context – one could say that signs become charged with meaning. Speech, for instance, owes this meaning-charge to something pre-linguistic. The meaning-charge is temporally restricted – it can be exhausted, as meaning can expire over time. For example, archaic words and symbols once charged with meaning are now sunken into oblivion, just because we do not use them anymore they have not been “re-charged”.

There are various branches of semiotics that argue that meaning and sign processes are ubiquitous. It might be important to point out that we do not ascribe to an interpretation of meaning as equivalent to processes of signification. Rather, as said, subjects provide the lived world with meaning prior to and without (the use of) signs. In fact, it is this very capacity that enables signs to become charged with meaning. Perception is, despite the lack of a sign structure (i.e. no differentiation between expression and content from the point of view of the subject), imbued with meaning [7]. To us, there is no question that ultimately there must be a “feeding channel” for meaning; channelling something that itself is external to the domain of sign use. Our analysis of meaning is at this level in concordance with a phenomenological approach: meaning in the first instance originates from an embodied and embedded fundamental first-person organizing of experience in the lived world.

To approach the organic ever-evolving sign system of language is not to approach something that has fallen out of the sky. Particularly, linguistic meaning is embedded in the culture and nature of human beings and it is preceded by perception and experience. It is founded, as Husserl would say. In the words of Zahavi:

To deny the existence of prelinguistic cognitive competencies, of prelinguistic syntheses of identification, and to claim that every apprehension of something as something presupposes language use not only makes it impossible to comprehend how the language user should ever have been able to learn a language in the first place, it also flies in the face of contemporary developmental psychology. […] for Husserl the analysis of the contribution of language is the analysis of something founded. (Zahavi 2003: 30)

We agree that linguistic meaning is founded in a “prelinguistic and prepredicative encounter with the world” (ibid.: 29). This is true in two ways, i.e. both in the historical/evolutionary/developmental sense and in the sense that experience always foregoes and conditions linguistic meaning. Taking the latter sense into account, we suggest that language should be seen as an instrument for meaning-sharing; the sharing of something that traces back to experience and intentionality. According to Husserl (ibid.: 22), the meaning of an experience is the intentional content (which he also calls the meaning intention). For our purposes we will change the nomenclature to intentional meaning. The intentional meaning of the act qualifies experience as to what type of intentional experience it is (e.g. recalling, emphasizing, imagining, fearing etc.), but more importantly, it also provides the aboutness of or directedness towards an object, i.e. the intentional object. On this definition, it is left open whether the intentional object exists or not, it is perfectly possible to be intentionally related to Kerberos. It might be useful to think of the intentional object as taking the object position in a grammatical subject-object relation. Moreover, an object in this sense can be everything from a keyboard to a mathematical equation, to a state-of-affairs – by no means is the understanding of objects restricted to physical ones, or any other certain kind. The way in which one is intentionally related to an object not only determines which object, but also the object as what. If I am holding a cup of hot tea, for example, the intentional content is that which both picks out which object I am experiencing (this cup of tea, rather than the moon or anything else) and as what I am experiencing it (as a cup of tea, rather than an alien potential danger or anything else). Thus, according to Husserl “we intend an object by meaning something about it” (ibid.: 23). We agree and mean that this sheds light on the fact that meaning traces back to the conscious organization of experience. Moreover, since we are from the very beginning in the world together, this fundamental organizing is necessarily made intersubjectively in a social context.

The intersubjective foundation of meaning

As shown, meaning tracks back to subjectivity and it should be clear by now that meaning is meaning for someone. When I perceive a chair-as-a-chair it constitutes a meaning act. Despite its lack of signified meaning, i.e. no differentiation between expression and content from my perspective, it still has intentional meaning. Prima facie, subjectivity could be just the vantage point of one particular conscious individual taken in isolation. It would however be wrong to analyze subjectivity without taking intersubjectivity into account. On our view, the ego presupposes the alter ego in its constitution [8]. To be a subject is inherently to be intersubjective. Thus, intersubjectivity and subjectivity are mutually entailing; the former is not just the coming together of two or more separated subjects – if that was sufficient then subjectivity would not necessarily involve intersubjectivity. Even if meaning is rooted in subjectivity, it makes no sense to speak of meaning without taking a second-person perspective into account. Just as I can perceive a chair-as-a-chair, so can you and I together perceive a chair-as-a-chair and partake in a dyadic act of intentional meaning in the absence of signs. As opposed to sharing signified meaning this is the moment of natural meaning sharing. The relation of profound recognition between the subject and the other is a necessary requirement for meaning to get off the ground and transcend the individual.

In interaction, there is a constant reciprocal action in the transmission of meaning connecting the intention of an intender with the interpretation of the interpreter. In other words, meaning, whether it is signified or intentional, is grounded in intersubjectivity. I intend something and you recognize my intention, and I recognize that you will interpret my intention. Here one might object that as I perceive e.g. a road sign there is no interaction between intention and interpretation – I am only interpreting a sign. Our reply is that there is an indirect interaction between the subject(s) intending something by putting up the sign and me interpreting it.

Relating our emphasis on intersubjectivity to advances in empirical sciences, we will now make a brief detour into some evolutionary and developmental psychological research that has received considerable attention within semiotics and linguistics (e.g. Sonesson 2006; Zlatev 2008). Sonesson (2006) points out that with some exaggeration one can speak of a “semiotic turn” within cognitive science. The research Sonesson has in mind takes “an interest in meaning as such, in particular as it has developed, ontogenetically and, in particular, phylogenetically” (ibid.: 138). Here, one finds, among others, the works of Donald (1991), Deacon (1997) and Tomasello (1999). Although there are substantial differences between their respective theories, one can at least find a clear connection in their insistence that culture (taken in a broad sense) plays an important role in the evolution of human beings. The prerequisites for the rise of culture, and together with culture, sign games, are largely given to intersubjectivity and highly developed cognitive and metacognitive skills that let us understand ourselves and others. Tomasello (ibid.) speaks at length about joint attention: the capacity to direct the attention of the other towards something by non-linguistic means. That is, I get you to see what I see and we both recognize that we direct our attention towards the same object from our respective vantage points. Similarly, Donald (1993) argues that explicit metacognition is crucial for intentional representation. The capacity to both represent an action in advance and review its success in retrospect is important for communicative (and intentional) gestures. Possibly, the capacity to represent the action of oneself is also applicable on our fellow beings. Thus, I can put myself in the position of what it would be like to have the kinds of experience that you have, and the same holds true for you. Even though these alleged capacities are representative to human beings, they are still in continuation with, and to some extent shared by, other social animals, e.g. great apes.

Leaving phylogeny at that, Piaget (1945; 1967; 1970) noted in his studies on ontogenetic development that between the age of 18 and 24 months children develop the capacity to — from the subjective point of view — differentiate expression and content. This capacity Piaget called the semiotic function. It should be noted that there is a certain ambiguity in Piaget’s use of “differentiation”. Sonesson notes that we can understand differentiation in two ways. It “may mean that expression does not continuously go over into content in time and/or space; or that expression and content are conceived as being of different nature” (Sonesson: forthcoming) [9]. Nevertheless, with the semiotic function, the child can let a given object stand for something else. Sonesson (ibid.: 152) notes the example of a child signifying candy with a pebble, thus having an understanding both of communicative intentions and how to produce signs. However, Piaget still argues that the experience of meaning ontogenetically antedates the semiotic function, but that it then does not suppose a differentiation between expression and content (Sonesson 1992). Just as with the example of the chair above: a subject can intend in the absence of making a differentiation between expression and content. It should then be noted that this function is regulated and facilitated by intersubjectivity and social interaction. Thus, for the pebble to signify properly the intention has to be in some way accessible and correctly interpreted by the interpreter — otherwise the child might try something else out as the signifier, or give up the whole project. The role of the interpreter in establishing meaning is in this simple example limited; the complexity of the intrinsic relation between intention and interpretation grows exponentially as the semiotic function plays out in complete sign games. Still, signifying intentions has to be directed from someone to somebody. Meaning is thus fundamentally interactive in its formation — involving both intended meaning and interpreted meaning. This dyad has to be taken as a whole — analyses of this interaction will not render either one any less indispensable. Obviously, this is not an argument for the notion that we are incapable of directing this process towards ourselves, or that intentions are impossible in the absence of an external interpreter.

To dig deeper into the topic of the evolution and development of human beings would go beyond the topic of this paper. Though these claims from the empirical sciences are in need of qualification, they still point towards the importance of intersubjectivity and features beyond signifying in the emergence of meaning. Consequently, the positions here presented point to conclusions in concordance with our argument proper: meaning is founded.

Without engaging on the issue of social cognition, we still want to state that it is certain that the capacity to deeply and directly recognize the other is crucial for the notion of meaning. In analogy with the notion of joint attention, we therefore wish to speak of a related claim, joint experience: when an experience is fundamentally shared, not only in the sense of via non-verbal means of directing the attention of the other. Joint experience occurs when an agent intends something with regard to the other, and the other recognizes that the agent intends something towards her and interprets this intention, and the agent recognizes that her intention gets interpreted by the other. If a subject can have experiences that constitute intentional meaning, we could speak of joint experience as facilitator of shared intentional meaning.

This taps in to the intersubjective constitution of the subject, meaning as shared intentional meaning is independent of sign games. To base shared intentional meaning in signified meaning would clearly violate the tenets of the phenomenological approach endorsed in this paper. Once we do have the capacity to signify we will of course make use of signs in all kinds of experience. So, as I point towards the chair, I direct your attention towards the chair and the pointing gesture might, at one level of explanation, correctly be said to be a sign. However, at a more fundamental level this is not rendered by the sign function as such – the shared intentional meaning of together perceiving the chair would be equally real in the absence of signification.

To delineate this basic capacity for meaning and meaning-sharing solely to the interaction between human beings would be to jump the gun. To what extent this is present in other animals (and in our interaction with them) as well we leave open to further investigation. What we wish to emphasize is that joint experience and a second-person perspective are necessary for meaning-sharing. Even more so, it does carry over to sign use. Recall Piaget’s semiotic function: as this capacity is realized, we make use of “semantic links” in the same sense as when sharing an experience. I can say “elephant”, or declaratively point, or draw a picture, and you will know (how to interpret) what I intend. Not only that, I will know that you know (how to interpret) what I mean. Through this, we can begin to share meaning and experiences that our interlocutor lacks. Shared intentional meaning can then be established via signified meaning. I can tell you about something that you have not experienced, or I can tell you something that you do not know. The wonder of how the linguistic sign (often far from iconic or indexical) can signify, seemingly with so minor an effort, is partly explained by the way it taps into our intersubjectivity and capacity to share experiences.

Now we have outlined the subjective and the intersubjective fundaments – to complete the analysis we have to take seriously the fact that meaning-sharing surpasses this dyad.

Consensus of meaning-sharing

When we speak of consensus, it should be taken as a notion wide enough not to make a sharp distinction between pragmatics and semantics. We believe this concept to be fairly compatible with Grice’s pragmatic analysis of meaning (Grice 1957); however, we do not consider pragmatics as an additional layer on top of a literal semantic reading (as in the case of speech act theory, or certain readings of Grice’s explication of ordinary language use). Being ironic, for instance, does not entail that pragmatic implicatures or illocutionary force profiles the meaning according to the intentions of the user — the means for irony is, rather, already encapsulated within consensus. Conventions and background information structures the environment of communication, denoting possible language and other sign use. The dimension of consensus takes us into the third-person realm.

Not only are we unable to strictly use a certain proposition in our own private way (as shown by the Private Language Argument (Wittgenstein 1953: §256-§271)); we cannot even use it in a private bilateral way. That is, even if we do — we do not, for as soon as we use the proposition, it becomes part of the system, it becomes official. Therefore we have to — at least partially — rely on the conventional use; in fact, when uttering the proposition we contribute to conventional use. One significant thing about this mechanism is that it allows us to speak about things of which we know nothing, still transmitting meaning, if only we have some understanding of how to use, say, a certain term. This supplies a division of labour that increases linguistic productivity and enables us to communicate knowledge and information that we ourselves do not possess, as observed by Putnam (1975). This is pretty obvious, what might not be obvious is that this very same mechanism forces us to conform to the intentions and interpretations of others.

In order to understand how meaning propagates through sign use, we must pay attention to the refined way in which contexts are determined sign-wise. They are determined so that they afford limited meaning-ranges (we will return to this concept). The basic setup of signifying communication that we have outlined in the previous sections – meaning intended and meaning interpreted – already specifies the possible meaning-ranges to a certain extent. An expression will always be influenced by the context; for example, when this intender says this to this interpreter in this situation there will immediately be produced a quite exclusive combination putting limitations on how the use of the expression can carry out meaning. On top of this, both intender and interpreter must make their respective judgments on the basis of consensus, cf. Figure 1 below.

Figure 1. The three dimensions of meaning

Sign games

Adopting Wittgensteinian terminology, consensus can be described as the general rulebook of language game management. However, for our purposes, we introduce the idea to extend the game analogy to include all types of signifying endeavours, thereby replacing “language games” with sign games. This notion may seem to be burdened by the analogy. The rules of most games are static whereas the rules, or maybe better, regulative functions, of the sign games must be thought of as conventional and dynamical (i.e. regulated by the activities of the sign game participants). The rules can change; or rather, they are bound to change in accordance with what the users are doing. As an example, language is constantly changing and we should consider the rules as far from absolute, rather rule of thumb-like. Sign game participants should then be conceived of as both rule-following and rule-shaping, both as regulated and regulating [10]. Furthermore, sign games will have to be regulated relative to other sign games, since they each bear on principles of use that breaks through into one another. If it seems that sign games are relatively isolated (the apparent major difference between the sign game of forming and testing a hypothesis, and the sign game of making a joke, for instance), this is only due to the fact that to survey the organic and tacit interplay between them often turns out to be too cumbersome a task. It is also important to distinguish between sign games and contexts – note the possibility of playing a particular sign game in different contexts; and further, the contextual factor may drastically influence the sign game in question, e.g. when making a joke at a funeral [11]. Another weight put on by the analogy in question is that whereas the rules of most games are fairly simple, the rules of sign games are undeniably complex. Participants have to pay attention to their shared perceptual experience in the present situation, the joint identification of context, norms and etiquette of sign use, as well as related social aspects; speech manner such as gestures, play of features, accentuation, tone of voice, etc. Still we accomplish this so effortlessly, even semi-automatically, in everyday life.

To recapitulate: there is no essential core in which the meaning of a sign is located such that it unites all the possible uses. Wittgenstein introduced the term “family resemblance” to account for what holds the “complicated network of similarities, overlapping and criss-crossing” (Wittgenstein 1953: §66) together. In our view language is characterized as the totality of language games that forms a — indeed very intricate — complex.

Meaning ranges

A theory that denies, as ours do, any reificationist view of meaning risks to collapse into meaning-nihilism. Our antagonist, the reificationist, might say that if one holds that there is no absolute meaning correlate to be found then there is really no meaning at all. And the antagonist is right in asserting that on our view one cannot sidestep vagueness, but that does not imply that there is no meaning at all. We argue that it is possible to steer a middle way between these two extremes. To show this, we will again stick to examples from natural language. Both intended and interpreted meaning is vague, as also noted by Wittgenstein (1953: §66). One way to handle this vagueness is to analyze it in terms of ranges. What motivates this move is that not only are there more often than not a number of different appropriate interpretations, but even intended meaning comes out unspecified. We seldom intend our utterances to be taken absolutely literally — actually it is plausible to question what that really would be like, if they were to be taken to mean something absolutely literally. What would the literal meaning of “I’m not sure that’s what they want us to believe” be? There is no strict literal sense to make of it, and the truth is that there is no strict literal sense to make of anything; the philosopher’s search for objective exactness has here been in vain. But that is not devastating to the effect of analysis. If we say, for example: “I think I need to go home now”, again, in a given context, what we mean is usually not something exactly; as a matter of fact, usually we do not mean anything exactly. We mean something within a range of meaning; we might mean “I will go home now” and/or “I will maybe go home now” and/or “I think I ought to go home now”. Nonetheless, this puts sufficient boundaries to speech-act contexts. For instance, by saying “I think I need to go home now” we might not mean “proceed to phase three and wait for further instructions”, unless, of course, in a very specific context, if it is a code for example. In other words, at the outset of an utterance there is a range of possible meanings, even for the intender. The way this works is that if there is a need for a more exact intention or interpretation of meaning, this will be specified by further expressions or other semiotic means. It would be a mistake to root analysis of meaning in language alone; yet, it would be an even worse mistake to root it in sentences or words. In the real world, meaning is mostly processed in dialogue, where it would not make much sense to speak of sentences or propositions as the entities that bear meaning. Often enough we do not even use complete sentences in the grammatical sense – speech, rather, comes in fragments. At this point, one might raise the objection that speech-acts and dialogue are only parts of language as a whole. Are we not forgetting about text? In reply we would eagerly abide by Saussure (1970 [1916]: 58): the signs of written language are a malformation. It is an incomplete formalization bearing completely on spoken language. Therefore, what has been said about spoken language also applies to text, only in a simplified way. Indeed, in writing we always have to compensate by being more specific, the same tendency is also mirrored in the rigorous rules of writing, as opposed to the relative freedom of speech [12].

Though we have mainly been focusing on language this idea affords expansion into other sign use as well. The same holds true for e.g. gestures. A pointing gesture of the hand might mean to direct attention, but of course there is ambiguity and context-dependency even in a simple case like that. In one situation it is an order, in another a warning of a danger closing in, in another it means “I will go there” or “You can’t go there” or even “Later we will put it there” and so forth.

The meaning-sharing network

Let us summarize our model of a meaning-carrying network: The basic element of organizing experience in first-person perspective, i.e. intentional meaning, determines preconditions of meaning converging intersubjectively in the form of joint experience. Meaning-sharing is comprised of two moments, which should not be understood as temporally sequential. The first moment of transmission is natural in that it is non-signifying and independent of a meaning carrying system. The second moment is the signifying one. It establishes semantic links between intender and interpreter by way of sign relations. Once a sign system is in place, the signs, now charged with meaning, develop into semantic links by virtue of mediating meaning between intender and interpreter allowing for meaning to propagate. Meaning is not exact but comes in ranges. The mediation of these meaning ranges in turn take patterns forming into sign games whose rules are both regulating and regulated by the participants in consensus, i.e. the dynamical totality of sign game regulating functions. Through joint experience and sign use we are inevitably and essentially connected to each other, and the overall nexus of this meaning broadcast forms what we now dub the Meaning-Sharing Network. The flow of meaning through the network is constantly shifting; in concordance with the members it represents the striving to organize and share experience, knowledge and information. The overarching structure of the network can now be set up as follows:

(1) The fundamental element of embodied and embedded    first-person organizing of experience in the lived world.

(2a) The natural meaning-sharing moment.

(2b) The moment of signified meaning-sharing.

(3) Consensus: The dynamical totality of sign game regulating functions.

This structure, we argue, is irreducible in the sense that every attempt to explain meaning by way of discriminating analysis, i.e. by concentrating exclusively on one of these domains, falls short of recognizing the living complexity of actual communication. It is easy to think of the meaning-sharing network as nothing but the natural language. But let us stress that natural language is only one of many types of signifying endeavours included in the network, language is just one of many instruments in the toolbox (in good company with the formal languages of logic and mathematics, symbols and non-linguistic signs, a great variety of social practices, art and body language).

Our main aim has been to argue that signified meaning has to be founded and to capture that meaning is neither, in Putnam’s famous words, “in the head” nor is it autonomous of the individual members that constitute it. Instead, we argue, meaning is in the network. A tempting metaphor for this living edifice is another familiar network: the internet. Just as it makes no sense asking where exactly a certain information segment is located on the internet, since it is spread and distributed, it makes no sense to ask exactly where meaning is situated since it is spread and distributed as well. This suggests that we are more profoundly linked together than we might think. Allegorically speaking, we are not islands, we are the ocean. [Ψ]


Deacon, Terrence 1997. The Symbolic Species. New York: W.W. Norton & Co.

Donald, Merlin 1993. Origins of the Modern Mind. Cambridge, MA: Harvard University Press.

Grice, Paul 1957. Meaning. The Philosophical Review 66: 377-88.

Itkonen, E. 2008. The central role of normativity in language and linguistics. In: Zlatev, Jordan, Timothy P. Racine, Chris Sinha and Esa Itkonen (eds.), The Shared Mind. Amsterdam: John Benjamins, 279–305.

Kripke, Saul 1982. Wittgenstein and the Private Language Argument. Cambridge, MA: Harvard University Press.

Piaget, Jean 1945. La Formation du Symbole Chez l’enfant. Neuchatel: Delachaux & Niestlé.

   — 1967. La Psychologie de I’intelligence. Paris: Armand Colin.

   — 1970. Epistemologie des Sciences de I’homme. Paris: Gallimard.

Priest, G. 2002. Beyond the Limits of Thought. Oxford, UK: Oxford University Press.

Putnam, Hillary 1975. The meaning of ‘meaning’. In: Gunderson, Keith (ed.), Language, Mind and Knowledge (in the series Minnesota Studies in the Philosophy of Science). Minneapolis: University of Minnesota Press, 131-193.

Quine, Willard van Orman 1960. Word and Object. Cambridge, MA: MIT Press.

Saussure, Ferdinand de 1970 [1916]. Kurs i Allmän Lingvistik (Swedish translation of Cours de Linguistique Generale). Budapest: Bo Cavefors Förlag.

Sonesson, Göran 1992. The semiotic function and the genesis of pictorial meaning. In: Tarasti, Eero (ed.), Center/Periphery in Representations and Institutions. Proceedings from the 3rd Annual Meeting and Congress of the International Semiotics Insitute, Imatra, Finland, July 16-21, 1990: International Semiotics Institute, Imatra, 211-256.

   — 2006. The meaning of meaning in biology and cognitive science: A semiotic reconstruction. Sign System Studies 34(1): 135-213.

   — Forthcoming. The mirror and the virtual world. From hominisation to digitalization. To appear in the acts of Segundo congreso mundial de semiótica y comunicacion: Monterrey, Mexico, Octubre 19 a 22, 2006.

Tomasello, Michael 1999. The Cultural Origins of Human Cognition. Cambridge, MA: Harvard University Press.

Wittgenstein, Ludwig 1953. Philosophical Investigations. Oxford, UK: Blackwell Publishing.

Zahavi, Dan 2003. Husserl’s Phenomenology. Stanford, CA: Stanford University Press.

Zlatev, Jordan 2008. From proto-mimesis to language: Evidence from primatology and social neuroscience. Journal of Physiology 102: 137-151.


[1] Frege’s well known theory, for instance, arguably positions itself to place second order concepts outside of the sphere of the effable whilst still speaking about them, thereby causing a contradiction. [Back]

[2] We do wish to stay neutral as to how one should go about in determining more exactly who qualifies as a member of the network in this sense. Needless to say, primatology has revealed at least a continuity in cognitive and social capacities between human beings and great apes. Refraining from taking an anthropocentric stand on these matters we will leave this question open.[Back]

[3] E.g. “the indeterminacy of translation”, Quine 1960. [Back]

[4] Cf. the interpretation of Wittgenstein presented in Kripke 1982. [Back]

[5] Whether language is similar to games in lacking a clear definition is another question. [Back]

[6] Please note that this discussion is not to be confused with the structuralist notion of arbitrariness; an analysis concerned with the internal configuration of the sign.[Back]

[7] This depends on our comprehension of sign. As stated in note 1, a sign is a sign if it stands for something for someone. Thus, a sign has to be something that is used by a subject to convey signified meaning.[Back]

[8] We wish to thank Oscar Ralsmark for providing us with this formulation. [Back]

[9] For elaborate discussions of the semiotic function, see the publications of Sonesson, e.g. Sonesson 2006. [Back]

[10] Thereby we see a path of combining the insights of the structuralistic motivation and its various counterparts. That is, the human structures are surely very influent but they stem from, and are upheld by, humans. [Back]

[11] Note that this is not to be confused with the distinction semantics/pragmatics; it is not the case that sign games are purely semantic whereas contexts are purely pragmatic. [Back]

[12] Please note that we surely recognize the distinctness and importance of written language with respect to e.g. the exogram function (Donald 1993), i.e. the capacity to “store” meaning over time. [Back]

[Ψ] We wish to express our gratitude to Andreas Lind, Joel Parthemore, Oscar Ralsmark and Jordan Zlatev for their insightful commentary on an earlier draft. We would also like to thank the editors of this issue for their helpful remarks. [Back]