No matter what activity we engage in, in order to achieve understanding some aspect of conversation and communicative activity is implicated and is absolutely essential. Communication assumes the logical pre-existence of signs. Semiosis, the process of using signs, provides a new notion of certainty and truth.


Iwould like to put forward a simple proposition, namely that conversation is foundational, but in an unusual way which has not generally been considered in debates on foundations. I offer the argument in this conference because propositions which assert that communication is foundational should have some relevance to communication scholars.

Propositions about absolute foundations or the nature of existence are traditionally dealt with in philosophy as part of metaphysics. So this paper is a contribution to the philosophy of metaphysics. In particular it is a contribution to a debate about absolute foundations that has been an issue in Western philosophy since the ancient Greeks. Briefly, the debate centres around the question of whether there is anything—in the world, in thinking, or ideas—about which we can be absolutely certain. Those who have asserted that there are things about which we can be certain are sometimes referred to as foundationalists (Bernstein, 1983). Against them some contemporary post-modernist thinkers have argued that there are no absolute certainties; among these we find pragmatists such as Rorty, deconstructionists such as Derrida, and philosophers of science such as Feyerabend.

The currently fashionable arguments against foundationalism proceed from a recognition of the central role that languages, discourses, and texts play in all our attempts to discover certainty. If certainty is dependent on these forms, the argument runs, we cannot escape into some external intellectual ‘space’, outside language where we can find certainty free of the local cultural and contextual features of language. Against this post-modern critique, foundationalists argue that unless there are some absolute foundations against which to judge knowledge, morality, the rationality of arguments, or the truth of propositions, then there is nothing left but relativism and subjectivity. This polarising is, of course typical of many of our intellectual debates. I would like to suggest a third possibility which, if correct, will allow both points of view to prevail, but in a rather different configuration.

For me the argument is not new: I have articulated the ramifications and consequences of the argument on many earlier occasions, but as it has always been as a supporting argument to other themes that were more central at the time, my heresy has gone unnoticed. It seems to me that there is now a case for bringing this particular argument into the foreground of debate.

I shall begin with the basic and central propositions and then spend the remainder of the paper exploring some of the ramifications and implications that flow from the central arguments.

Communication is essential

Put simply, the argument runs like this. Whatever argument one advances, for whatever purpose, one thing is absolutely essential: the presence of argument. If one wishes to articulate one’s thinking, no matter what the thoughts may be, there is one thing which is absolutely essential: the process of articulation. If one wishes to engage in conversation, no matter what the subject, one thing is absolutely essential: the conversation. If one wishes to write a treatise on philosophy, even a treatise on the end of philosophy, one thing is absolutely essential: the writing.

No matter what activity we engage in, in order to gain knowledge or understanding some aspect of communicative activity is implicated and is absolutely essential. Even if you would like to refute this proposition, you can only do so by argument, articulation, conversation, or writing. In other words, communication is in an important sense foundational.

There are two ways of dealing with this central and incontrovertible proposition: one can either accept it and explore its profound implications, or one can say that it is only trivially true. Clearly I would like to assert the former. But I shall deal first with the argument that says the proposition is only trivially true.

In In Search of Semiotics, I dealt with the argument by focusing on the search for foundations in mathematics and logic as follows.

Some philosophers have argued that logic is the basis of any kind of reasoning, and since philosophy is, above all else, concerned with reasoning, logic underlies philosophy. This has led them to look for fundamental principles in logic as a basis for our conception of reality. However, logic is not only dependent on signs, it is inconceivable without them, by which I mean that we could not even imagine the possibility of logic without signs. Thus the process of making signs—semiosis—is fundamental to logic; semiosis is protological.

However, this may seem a weak argument. If signs were merely the vehicle for expressing ideas, then the primacy of semiosis would only be trivially true. The ideas themselves would be far more important. Like people on a train, the ideas are more important than the means by which they travel. We would not want to argue that the existence of the people is dependent on the prior existence—either logical or chronological—of the train. It may therefore seem that the claim on behalf of semiosis is hollow. But the simile of the train or vehicle is simply another expression of the idea of communication as transmission.

And here is the core of the argument. Traditional philosophers have been able to assert the argument as trivial because they have either implicitly or explicitly accepted the transmission view of communication. They have been able to avoid dealing with the central role of texts or discourse in philosophy by assuming that these were merely the apparatus of transmission, not the substance of ideas, arguments, or evidence. The core concerns of philosophy, according to traditional philosophy, remain outside language in the realms of thought, reason, or the world. Philosophical discourse—the use of language—is merely the vehicle through which these profound issues are to be discussed.

There are now substantial arguments against this transmission view of communication and I shall not bore you with recycling them in this paper (see Reddy 1979, Sless 1986).

For those of you unfamiliar with these arguments, but familiar with the conventional texts in communication studies, I should point out, in parentheses, that the arguments against the transmission view are not the flimsy rationalisations that are to be found in such texts as Fiske’s Introduction to Communication, which falsely contrasts the transmission view with the semiotic view of communication. Careful analysis reveals that Fiske’s semiotic view, far from countering the transmission view, simply takes it for granted. Indeed, semiotics as found in cultural studies, with its use of the encoding/decoding schema, is transparently a variant of the transmission view of communication. The Fiske arguments may be a legitimate ideological attack on a particular empiricist approach to communication research. But these arguments have nothing to do with philosophical issues about the nature of communication. On philosophical grounds there is nothing to choose between either the semioticians or the empiricists. They both agree that communication is a process of transmission. The only difference is one of emphasis: the empiricists seek to explore the causal relation between sender and receiver; the semioticians presume the causal relation, and seek to explore how meaning is created within the process.

The arguments against the transmission view show that communication as a means of conveying ideas from one point to another is false. In contrast, the new view of communication argues that ideas emerge from the communication process. This leaves the traditional philosopher unable to assert that my starting proposition is only trivially true. The philosopher who wishes to assert the importance of ideas must take special account of communication since it is in the communication process that ideas are formed.

Thus we can now turn from a consideration of communication as only trivially foundational to the argument that communication is profoundly so.

Post-modern philosophers such as Derrida, Foucault or Rorty have understood the profound relationship between communication and reflection and have correctly asserted that the texts play a central role. But, as I shall show, they have wrongly concluded that this means the end to metaphysical or foundational claims. They have, as it were, thrown out the baby with the bathwater.

A digression into history

To understand the fallacy in their reasoning, it is necessary to digress into the history of philosophy. Only when we understand the historical background to contemporary debates can we see why particular false conclusions have been drawn.

The post-modern philosophers stand in opposition to the classical Western tradition, starting with the Greeks and extending to modern philosophers such as Descartes, which tries to deal with metaphysical issues in three ways: to use the human mind as the central axis for discussions of the nature of existence; to appeal to some notions of pure reasoning; or finally to appeal to experience or scientific observation. None of these approaches or their hybrids have been totally successful, and post-modern philosophers argue that they have all foundered on the rock of language.

This has been the profound insight of twentieth century philosophy. Wittgenstein more than any other philosopher pointed to this in his seminal arguments about ‘language games’. The essential feature of the argument is that we cannot escape the pervasive fabric of language in a search for traditional notions of truth.

The post-modernists are correct in asserting that this realisation brings to a close a long chapter in the history of philosophy which began with the ancient Greeks—a two and a half thousand year quest for a particular kind of ultimate truth.

But I would like to emphasise that the quest by some Greek philosophers, especially Plato and Aristotle, was for a particular kind of ultimate truth. There were other Greek philosophers, in particular Heracleitus, who from our position of hindsight might be understood as pursuing very different notions of truth. And it may well be that in the debris of post-modernism, certainty and absolute truth still have a place if we return to these other Greek thinkers.

A useful point of departure, or more accurately a point of beginning a new type of debate, is to look at a moment in the history of philosophy when we can see the argument about certainty and truth take a particular direction. I do not make any historical claim about this moment; I have chosen it for what it shows us today, rather than what it may tell us about choices made at the time.

The point of beginning I wish to take is the moment in Plato’s Theaetetus when he discusses and dismisses the possibility of knowledge from experience based on the Heracleitean view that everything is in flux. Notice how powerfully language is implicated in the dismissal.

…. It seems that what we have actually discovered is that, if all things are in process of change, any answer to any question whatever is equally right: one may say it is so and it is not so—or ‘becomes’, if you wish to avoid a term that might bring our opponents to a standstill.

…. Except…. that I used the words ‘so’ and ‘not so’. This ‘so’ is inadmissible: a thing which is ‘so’ would no longer be changing and the same applies to ‘not so’. Some new jargon will have to be devised for those who teach the Heracleitean theory, for at present they have no terms in which to cast their fundamental propositions—unless perhaps ‘not even no-how’! That might be sufficiently indefinite for their purpose. (Plato, trans Warrington, 1961, pp117-8)

We can, of course, only speculate on what might have happened if a different direction had been taken, if the ‘jargon’ which Plato mockingly alludes to had indeed existed. But from our vantage point two thousand five hundred years later we can make two observations. Firstly, Plato treats language as a vehicle, a container ‘in which to cast … fundamental propositions’. This makes it possible for words to be the stable carriers, the means of reliably transmitting ideas. But if words correspond invariantly to certain ideas, neither words nor ideas can change. Secondly, and more fundamentally, Plato cannot come to terms with the idea of continual change. He struggles with the idea when he suggests the term ‘becomes’, but goes no further. The problem that confronted Plato, and which has confronted generations of philosophers since, is that they find it difficult to conceive of order and change simultaneously. These two notions have seemed antithetical. Change has been seen as synonymous with a lack of order—a swamp of chaos, irrationality and madness—something feared by philosophers; what Bernstein refers to as the Cartesian anxiety, but which, quite clearly from a reading of Plato, has a much earlier provenance. Yet I believe that there is a productive way of reading Heracleitean theory which does not lead in this direction.

One of the most commonly-used metaphors to characterise the Heracleitean theory was the idea of a river.

Heracleitus somewhere says that all things are in process and nothing stays still, and likening existing things to the stream of a river he says that you would not step twice into the same river. (Plato Cratylus 402A, from Kirk & Raven 1964, p. 197)

As we saw from the Theaetetus, this suggested to Plato an unacceptable irrational outcome in which any answer to any question could be equally right; a clear resonance with our contemporary concern with the problems of relativism and subjectivism. But if we look at a fragment of a manuscript, thought to be closer than Plato’s account to Heracleitus’s original writing, we can read the river metaphor in a different way.

Upon those that step into the same rivers different and different waters flow… It scatters and …gathers…it comes together and flows away…approaches and departs. (Kirk & Raven 1964, p. 196)

This is not the river which is never the same river twice. On the contrary, it is a river in which the change is orderly, regular, and the same, even though the substance—the water—is different. This is a celebration of the orderliness of things, not an invocation of disorder. But it is the orderliness and stability of change, not the stability of ideas or their names which Plato sought. None the less it is true that there was no ‘jargon’ in Plato’s day that could articulate such change.

Our culture had to wait two thousand years for Newton’s differential calculus before we had our first powerful means of describing change through an orderly set of mathematical equations.

Since then we have had many mathematical systems giving models for articulating change. The most obvious modern ones are catastrophe theory, which provides an elegant mathematical description of catastrophic changes of states; and chaos theory, which suggests that chaotic turbulence, such as that found in the water that so alarmed Plato, may well be the product of underlying regularity. Order and change are therefore not antithetical, even though it may have seemed so to the Socratic mind.

Yet our contemporary philosophers see no further than the old ways and the old notions of certainty that are so earnestly debated in the Socratic dialogues. Even when, as with the post-moderns, they dismiss the old notions of certainty, they do not see, or perhaps wish to see, the possibility of a more powerful order and certainty waiting in the wings. So for example Rorty claims that:

One can use language to criticise and enlarge itself, as one can exercise one’s body to develop and strengthen and enlarge it, but one cannot see language-as-a-whole in relation to something else to which it applies, or for which it is a means to an end. The arts and the sciences, and philosophy as their self-reflection and integration, constitute such a process of enlargement and strengthening. But Philosophy, the attempt to say ‘how language relates to the world’ by saying what makes certain sentences true, or certain actions or attitudes good or rational, is, on this view impossible.

It is the impossible attempt to step outside our skins—the traditions, linguistic and other, within which we do our thinking and self-criticism—and compare ourselves with something absolute. (Rorty 1987, p. 33)

Can the ubiquity of language ever really be taken seriously? Can we see ourselves as never encountering reality except under a chosen description…? (ibid p. 57)

While the substance of Rorty’s argument is different from Plato’s, the form is not. Rorty is here asking the same question as Plato and coming to the same conclusion, though unlike Plato, Rorty does not fear relativism.

In our time we have come to realise that all discourses are potentially subject to multiple interpretations. Language, like the waters of the river, is in constant turbulent change. And Rorty, though taking the opposite view, repeats the form of Plato’s argument with today’s concerns: in the river of language all is in a process of constant change, and any answer to any question whatever is equally right.

Plato trusted certain formal types of language in a way that we don’t. Further, he assumed that certainty and truth could be discovered independently of language, but we do not. Therefore the route Plato took, in developing notions of certainty and truth away from the unpredictable turbulence, is for us blocked. It is perhaps ironic that Bernstein, in an attempt to provide a post-modern synthesis which does not lead to absolute truths, reinvokes something akin to the idyllic Socratic dialogues as the way ahead.

In all of them (Gadamer, Habermas, Rorty, and Ardendt) we have felt a current that keeps drawing us to the central themes of dialogue, conversation, undistorted communication, communal judgement, and the type of rational wooing that can take place when individuals confront each other as equals and participants. (Bernstein 1983, p. 223)

Unfortunately the Socratic dialogues took place against a background of gross inequality, and the dialogues themselves were masterpieces of stage-managed intellectual dominance by a particular point of view. In our own time, as then, intellectual debates are inextricably tied to issues of power and inequality. ‘Rational wooing by equals’ requires the prior existence of equality before the conversation can begin. And it may well be that in a climate of equality, there would be little to discuss or argue about.

But the alternative route, which neither Plato, Rorty, or Bernstein considered, is still open. It does not necessarily follow that because all things are in change there is no order, no certainty. Even Rorty implicitly accepts this when he exhorts us to keep the conversation going. It is precisely in the mapping of the dynamics of conversation that our certainties lie.

Certainty in semiosis

I shall now move to a demonstration of the existence of these certainties. But first a formal point. The method of argument I use is based on the principle of parsimony, or Occam’s razor: to use the simplest set of ideas necessary to sustain an argument. In adopting such a method I have not discounted or ignored the obvious and massive complexity of communicative experience, its richness and unpredictability. Economy should not be confused with reductionism. I do not believe that complexity can be reduced by simple propositions. Rather I have taken a particular intellectual style of argument in the belief that a rigorous and economical set of propositions and arguments will serve best in the long run in advancing our understanding. I believe in economy, simplicity, and elegance as a guiding aesthetic in the creation of theory. This belief has been stubbornly maintained during a period in which debates about communication have taken on an almost gargantuan complexity, with many of the arguments obscured by mystification. With the passage of time and the passing of intellectual fashions, I am now more than ever convinced that the course I have taken is the correct one.

I started with the proposition that no matter what activity we engage in, in order to gain knowledge or understanding some aspect of communicative activity is implicated and is absolutely essential. From there we can make an observation that must necessarily be true of all communicative phenomena: communication depends on signs. All notions of communication assume the logical pre-existence of signs. Leaving to one side, just for the moment, any precise definition of what a sign is, but accepting a broad notion of signs as having a representational function, we can assert that without signs communication could not take place.

For example, without language, conversation would be impossible; without the alphabet, writing would be impossible. Putting the issue negatively, there is no case where communication could take place which does not depend on signs. There is some important and irreducible sense in which communication is dependent on signs.

It follows from this observation that we need to have some notions of signs in order to carry the argument further.

There are two traditions of thinking about signs in common use today: one derives from Peirce and the other from Saussure. My own approach derives more obviously from Peirce, though in many important respects, which fall outside the purview of this paper, there are substantial differences between Peirce’s view and my own. Many of the detailed arguments for not using Saussure as a basis are contained in In Search of Semiotics (Sless 1986, referred to here as ISS) and there is no need to rehearse them here. For those interested in a more thorough rejection of Saussure’s thought I suggest Roy Harris’s excellent work (Harris 1986). Suffice it to say firstly, that Saussure’s notion of a sign, compared to Peirce’s, generates such a degree of contingent complexity, and such an assortment of unresolved issues, as to disqualify it on the grounds of parsimony; and secondly, the difference between Peirce’s and Saussure’s notion of a sign is so substantial as to make any reconciliation between them impossible. Peirce’s notion (though not his elaboration of it) is philosophical, while Saussure is concerned entirely with a quasi-psychological process. This critical difference between the two is at times glossed over (Fiske 1982). A close reading of their respective works—each generated in ignorance of the other and in radically different cultural settings—reveals deep and irreconcilable differences.

Peirce offers the following definition of a sign:

A sign is something which stands to somebody for something in some respect or capacity (Peirce 1931–58)

A sign consists of a tripartite relation between a sign, a user, and what the sign stands for (called here a referent because it is what the sign refers to). These three elements are inseparable in the sense that something can only be a sign if it is involved in this tripartite relation.

There is a need to say something about each of these elements in turn as I have defined them in ISS.

Users in the sense in which I have employed the term can be either collective or individual—classical individuals in the Cartesian sense, or subjects inscribed within a discourse in the Foucaultian sense. The type of user, invoked in any particular application of the concept of sign, does not affect the definition of the sign.

Signs, as the definition implies, are only signs by virtue of the use made of them. It is meaningless to talk about the properties of a sign on its own, since a sign, qua sign, has its properties only by virtue of the relation it is in. In principle anything can be a sign. A sign can be a material object, imaginary object, idea, even another sign. It derives its sign properties by virtue of its relation to users and referents, not because of anything intrinsic to it. The only partially intrinsic property of a sign is that it should be distinguishable by the user from its referent. This distinction is necessary because a sign cannot stand for itself.

Similarly, a referent can be a material object, imaginary object, idea, even another sign, and it enjoys its properties by virtue of the relation it is in with a user and a sign.

The boundaries of signs, referents, and users are not necessarily sharp, though this is often assumed by other contemporary theorists. For example, Chomsky has claimed that one of the distinctive characteristics of words is that they are finite entities with sharp boundaries between one word and the next. But as Illich and Sanders (1988) show, even the concept of a word is of relatively recent historical origin, an invention of literate societies. Further, if one regards a word as a sign within a triadic relation, not just as in a relation with other signs, there are no necessary sharp boundaries. As Wittgenstein showed, a word such as ‘game’ has a series of vaguely overlapping uses (Wittgenstein 1953). Only one aspect of the characteristic of the triad ‘game’ is finite; as the users change so might the referent, and any particular user can have either a vague of clear sense of the referent they are deploying. Signs, referents, or users can be vague entities or sharply delineated ones. Nothing in the definition I have given of a sign necessarily implies either a sharp or vague delineation.

The stand-for relation

The engine that drives the relation between sign, referent, and user is what I have described in ISS as the stand-for relation. The process of linking signs and referents through the stand-for relation is called semiosis. In my definition of the sign, the stand-for relation is invoked by the user and acts as the link between the sign and its referent. Through the stand-for relation we can not only account for fundamental processes of understanding but also crucially locate agency within the process.

Other theories of semiosis do not recognise that there is something here which requires explanation. Simply to assert, as Saussure does, that there is an associative link between a signifier and a signified does not tell us how this link comes about. What is the process by which these things come together? Even Peirce, from whom my own thinking derives, gives an unsatisfactory account of this central process.

Peirce’s explanation of how a sign works is as follows:

It [the sign] addresses somebody, that is, creates in the mind of that person an equivalent sign, or perhaps a more developed sign. (Peirce 1931–58)

This leaves us with three problems. Firstly, the sign is said to ‘address somebody’. This implies, if we take it literally, that the sign can act as if with its own volition. A theory of animistic action would strain any contemporary theory of communication, let alone one concerned with economy of explanation! Secondly, Peirce says that the action of the sign ‘creates in the mind of that person an equivalent sign’. This leaves us with the problem of accounting for how this creation takes place in psychological terms. The answer that many have given, particularly those who have followed Peirce, such as Morris, is that the relationship is the mechanistic one of stimulus and response. Finally, anticipating behaviourism, Peirce says that it is the sign rather than the user of the sign which ‘creates’ a response. The user can merely react, never act. This leaves out the question of how new signs are created.

Letness and the origins of signs

When we come to this most basic of questions, how new signs are created, we come to the central issue of this paper and what I believe to be foundational in communication.

When I tackled this question in ISS, I did so through a discussion of Gödel’s theorem—a mathematical proof that undermined the search for foundations in logic and mathematics. (For a technical discussion of the argument, reference to ISS would be worth while.) In this paper I will quote from ISS to give a sense of the argument.

Every mathematical system depends on statements which take the form ‘let x stand for y’. Once x is given its new status, the operations performed on it are as if it were y…However, ‘let x stand for y’ is not part of any rule inside a mathematical system. It is the method by which the system comes into existence. The axioms of mathematics come from such humble primary propositions. ‘Let x stand for y’ is the simplest expression of the nature of semiosis. The core operation which links the x and the y is contained in the term ‘let’.
….We need to speak of this basic quality of semiosis in a special way. I shall allow myself the luxury of one neologism—a term that will identify the core of semiosis and allow us to understand its nature; I shall call it letness.

Letness is characterised by a fundamental anarchy. It is subject to no logic, no rules of inference, no causal relations or moral imperatives. We may of course attach these things to letness retrospectively or even at the time when a new stand-for relation is created but there is no necessary requirement for letness to be subject to any imperative. Further, letness is not reducible to some other state, condition or explanation. When a mathematician says ‘let x stand for y’, we cannot reduce this statement down to some more basic construction—untie its logical knots or reveal its inner workings. It stands alone. I take letness to be the central metaphysical necessity of the semiotic point of view.

I intimated with my return to Heracleitean theory that any new notion of certainty would be dynamic, by which I meant that we would not find certainty in things or ideas—what Derrida referred to as the metaphysics of presence—but rather in the processes by which ideas or things come to be known, not known, understood, misunderstood and changed. At the heart of that process is what I have described in ISS as letness.

I went on in the preface of ISS (which was of course written after the book was finished) to say something about how understanding—the commodity traditional philosophy most sought after—fared under this new metaphysics.

[A]llow me to caution you about the nature of understanding. Understanding is achieved when, for a moment, there are no more questions to ask. Understanding is the dead spot in our struggle for meaning; it is the momentary pause, the stillness before incomprehension continues; it is the brief relief from doubt that is the norm. Thus understanding is a temporary state of closure. When we understand something we are effectively saying there is no more to ask, no more to question, all is revealed. But of course ‘all’ is never revealed and the sensation of certainty always passes.

Perhaps somewhat poignantly this shows how the classical search for certainty and truth was always doomed to fail; the certainties and truths about semiosis make it inevitably so. The impossibility of classical truth is demonstrated by what is true of semiosis, and by extension all communicative processes.

Where does this leave us? Again I quote from ISS:

The landscape of communication is more like the surface of a giant trampoline than terra firma. When a trampoline yields as we walk across it the feeling may be one of uncontrollable and hence chaotic movement but we know that the trampoline is obeying strict physical laws of elasticity which do not change. The regularity is simply at a level which as walkers we have not yet grasped. In shifting from a metaphor of a landscape to that of a giant trampoline I am trying to convey a sense of the level at which order and perhaps truth is to be found in semiotics.

One final thing that bears repeating in the current debate is that we keep on writing, talking and arguing, even as we say there is no firm ground underneath us. This simple fact leads me inescapably, as I hope I have demonstrated, to the conclusion that a particular kind of certainty is to be found in the very fabric of communication. It is not the specifics of particular language games that provide the basis of philosophic certainty—that was the mistake of classical philosophy with its cultural centricism—but the absolute fact and necessity of language games which provides the basis of philosophic certainty. It is in and through communication that we can see a pattern which is certain.


Bernstein, R. J. (1983) Beyond Objectivism and Relativism, Oxford: Basil Blackwell.

Fiske, J. (1982) Introduction to Communication Studies, London: Methuen

Harris, R. (1986) Rereading Saussure, Oxford: Basil Blackwell.

Illich, I & Sanders, B. (1988) The Alphebatization of the Popular Mind, London: Marion Boyars.

Kirk, G.S. & Raven, J.E. (1964) The Presocratic Philosophers, Cambridge: Cambridge University Press.

Peirce, C S. (1931-58) Collected Papers, Cambridge Mass: Harvard University Press.

Plato (1961) Parmenides and Other Dialogues, translated by John Warrington, London: J.M.Dent and Sons Ltd.

Reddy, M.J. (1979) The Conduit Metaphor, in Ortony, A. (ed) Metaphor and Thought. Cambridge Mass.: MIT Press.

Rorty, R. (1982) Pragmatism and Philosophy, in After Philosophy (1987), edited by K. Baynes, J. Bohman, and T Mc Carthy, Cambridge Mass: MIT Press.

Sless, D. (1986) In Search of Semiotics, London: Croom Helm.

Witgenstein, L. (1953) Philosophical Investigations, translated by G. E. M. Anscombe, Oxford: Basil Blackwell.