Okay, having covered Eco’s introduction I will now summarise his first chapter. Of the chapters I’ve read so far (I’m about two thirds of the way through) this one was the most difficult to understand the thought progression so I’ll try to break it down as clearly as possible with the hope that I get what Eco is talking about.
[1] Signs
Quick Overview: Because this is a book of semiotics (i.e. the study of signs), Eco begins at the beginning and explains what exactly he understands a ‘sign’ to be. He begins with traditional definitions of signs and explains why such definitions have been called into question [1.1]. He then goes on to summarise several ways in which dictionaries and everyday language talk about ‘sign’ [1.2], and then asks if a sign must be intensional or extensional [1.3]. At this point he examines those who propose that signs should be understood as a linguistic entity [1.4], before looking at the multiple ways in which contemporary scholars have deconstructed the linguistic sign [1.5]. This discussion leads Eco to reexamine the relationship between signs and words [1.6] and he examines how the Greek Stoics began to formulate a solution [1.7] which was then developed by Augustine and later models that unified linguistic theories and semiotic theories [1.8]. It is this approach to signs that has been so vigorously attacked by the scholars in [1.5], and so, in response to this Eco begins to develop his own “instructional model” of signs, which overcomes the dichotomy often posited between signs and semiotics [1.9]. In support of his theory, Eco discusses “strong codes and weak codes [1.10], and looks at abduction and the inferential nature of signs [1.11]. This leads him to his criterion of interpretability, which is basically the thesis of his book [1.12], and he then closes with a few remarks on the relation between sign and subject, which prepares the way for chapter 2 [1.13].
Okay, so that’s all well and good but what the hell does this all mean? I’ll break it down, section be section.
In [1.1], Eco mentions traditional definitions of ‘sign’ which see a sign as “something which stands to somebody for something in some respect or capacity” (Peirce). This is basically a restatement of the classical definitions which understood a sign to be aliquid stat pro aliquo (i.e. “that which is there stands for that which is not”). Thus, as Saussure suggests, a sign is a twofold entity, it is a signifier and a signified. Or, as it is more technically put by Hjelmslev, a sign is a mutual correlation between two functives, the expression-plane (i.e. signifier) and the content-plane (i.e. signified). However, these definitions are undergoing a crisis, and some have thus proclaimed the death of the sign. This is so for a few reasons. The first is because of how ‘sign’ is commonly used, and the second is because of a sustained assault be scholars upon these traditional definitions.
Thus, in [1.2], Eco examines many ways in which dictionaries and everyday language talk about ‘sign’.
The first way sees the sign as a manifest indication from which inferences can be made about something latent. This can occur in two ways: through a synechdochic relationship (i.e. as if the sign were a peripheral manifestation of something which does not appear in its entirety) or through a metonymic relationship (i.e. as if the sign is a visible imprint left by an imprinter on a surface, and the imprint tells us something about the imprinter). The key to this type is an inferential mechanism: it argues that if p then q.
The second way sees signs as a gesture produced with the intention of communication. Here a relation of equivalence is important, not a relation of inference. Thus p is equivalent to q.
The third way sees signs as symbols which represent abstract objects, like logical formulas or diagrams. There are also inferential and iconic or analogical.
The fourth way sees signs as diagrams, or pictures that bear a likeness to that which they represent.
The fifth way sees signs as drawing in an abstract or stylised form. These are also iconic but are arbitrary because they are in a state of catachresis (i.e. the are used in ways that violate the norms of a language community; they are used in the ‘wrong’ context).
The sixth way sees signs as an expression where the aliquid instead of standing for, stands where a certain operation is to be addressed. So this type is instruction. If p now, and if you therefore do z, then you will obtain q.
Therefore, in [1.3], Eco argues that we clearly have too many things that are ‘signs’. In order to overcome this problem we need to ask if signs are an intensional device (i.e. a set of all possible things a word could describe) or an extensional device (i.e. a set of all the actual things a word describes). Are we looking at relationships of equivalence or of inference? How far can this inference be pushed? And what constitutes a sign? The actual phonetic utterance of a word or the lexical model which the word represents?
Eco begins to examine some “elusive solutions” in [1.4]. First, he points to people like Malmberg who argue that the word ‘sign’ only applies to linguistic entities. ‘Symbols’ are these elements which represent something else, while ‘signs’ are reserved for language. However, Eco argues that this approach does not (a) explain to what extent signs are related to symbols; (b) which science should study symbols and which categories should be employed; and (c) the relationship between intension and extension is not clear. Thus, this approach is too problematic.
Second, Eco points to scholars like Harman who break down the traditional definition of a ‘sign’ into three separate theoretical subjects: (1) the theory of the intended meaning; (2) the theory of evidence; and (3) the theory of pictorial depiction. However, this approach is also too problematic because it clashes with the linguistic usage of ‘sign’ which has united these three categories for over 2000 years, and it also clashes with the ‘philosophical instinct’, which realises how inextricably linked intended meaning is with evidence and pictorial depiction.
In [1.5], Eco then goes on to examine further attempts to deconstruct the linguistic sign. Apart from a focus on linguistic signs, all these attempts tend to dissolve the sign into entities of greater or lesser purport.
First, he examines Hjelmslev’s approach which distinguishes between signs and figurae. Hjelmslev sees the traditional sign as too large and so he identifies figurae at the content level. Thus, language is not a pure sign system. Language is a system of figurae that are used to construct signs. However, this approach is problematic because it does not account for signs that cannot be broken down into multiple figurae.
Second, scholars like Buyssens argue that the sign is to small and the semantic unit is the seme (i.e. the sentence). This shifts the focus from language as a system of signification to language as a process of communication, but it is problematic because it divides the two approaches when they should be understood as complementary.
Third, Saussure argues that the elements of the signifier and of the signified are set in a system of oppositions in which there are only differences. The correlation of the content-plane with the expression-plane is also given by a difference. Thus the sign exists by a dialectic of presence and absence. Thus, along the lines of Derrida and (much earlier) Leibniz, the whole sign system dissolves into a network of fractures, and since the presence of one element requires the absence of another the argument is “autophagous” (i.e. self-devouring).
Fourth, Lacan argues that the semiotic chain is merely a chain of signifiers. The signified is continually transformed into a signifier through a process of perpetual substitution. However, this is merely a misunderstanding based on wordplay for, in any given context, there will be a specific signifier, and a specific signified.
Fifth, there is the view of Barthes, Derrida, and Kristeva, that argues that signification occurs exclusively within a given text. Thus, meaning is located solely within a text, and the text alone determines the meaning of a sign. However, this is problematic because texts are always produced within communities and the draw from, and manipulate, signs that pre-existed the text.
Sixth, Kristeva further develops the notion of a sign based on categories of ‘similitude’ and ‘identity’. This is rooted in the notion of equivalence. However, this too is problematic for the equivalence model is a degenerate notion of linguistic signs.
Thus, having highlighted the various assaults on the traditional linguistic ‘sign’, Eco steps back to trace the development of the notion of the linguistic sign, in order to show how these critiques gained plausibility, and in order to then provide his alternative.
In [1.6], Eco looks at the relationship between the ‘sign’ and the ‘word’, beginning with Hippocrates and Parmenides. Initially the term ‘sign’ was not applied to words. Words were names, deceptive representations imposed on objects we think we know, that establish a pseudoequivalence with reality. ‘Signs’ were seen as evidence, inferences about the nature of reality. However, Plato and Aristotle shifted the focus of the discussion and words were examined looking at (a) the difference between the signifier and the signified; and (b) the difference between signification and reference. Signification says what a thing is, reference says that a thing is. Throughout Aristotle’s writings there is a reluctance to use the term sign (semeion) for words; rather, words were seen as symbols, a more neutral term at that time. Aristotle continues to see signs as proofs/evidences, and as such they only apply to words when he talks about words are a proof that one has something in one’s mind to express. He does not equate linguistic symbols with natural signs.
However, in [1.7], Eco examines how the Stoics propel the issue forward. The Stoics do not totally integrate a theory of language and a theory of sign. What they do is create a distinction between expression, content and referent. The expression is the actual articulated sound, the referent is that which the sound refers to, and the content is an incorporeal bridge between the two. It is a lekton (i.e. an ‘expressible’). Signs refer to something immediately evident which leads to some conclusion about the existence of something not immediately evident. As such signs can be commemorative deriving from an association (confirmed by preceding experience) between two events, or they can be indicative pointing to something which has never been evident, and probably never will be. Therefore, for the Stoics the theory of language becomes closely associated with the theory of signs. Signs only emerge insofar as they are rationally expressible through the elements of language, language is articulate inasmuch as it expresses meaningful events, and language thus becomes the primary system through which all other systems are expressed.
Thus in [1.8], Eco looks at how the theory of signs and the theory of language are unified, with the predominance of linguistics. Augustine unites the two theories arguing that signs is the genus while linguistic signs are the species, but overtime the model of the linguistic sign gains dominance and is seen as the semiotic model par excellence. Thus The linguistic model was flattened and crystallised into the form used by the dictionary and by a lot of formal logic which had to fill its empty symbols only for the sake of exemplification. Therefore, the notion of meaning as synonymy gained precedence and lead us to the situation in which we now find ourselves.
Therefore, in [1.9], Eco proposes a way forward — “the instructional model” — which differs from the critiques raised by the deconstructionists. Eco’s solution is that the meaning of a categorematic term (i.e. a term or phrase capable of standing as the subject or [especially] predicate of a proposition) or syncategorematic term (i.e. a term or phrase that cannot stand as the subject or [especially] the predicate of a proposition) is found in a set (a series, a system) of instructions for its possible contextual insertions and for its different semantic outputs in different contexts (all registered by the code). Thus, in this approach the semantic type is the description of the contexts in which the term can be expected to occur. This takes place through inference, and the process of recognising natural-events that generate sign-propositions also relies on this inference. Therefore, there is no difference between first level signification (i.e. signs) and second level signification (i.e. semiotics). The sign only exists because is has been constructed by the discipline which studies it.
Okay, so the meaning of a sign is produced through a contextual code but what is this code? In section [1.10], Eco examines strong and weak codes. This is basically an examination of the epistemological value of if and then within any context. Here Eco draws on Aristotle’s theory of necessary signs and weak signs, and this is tied to the notion of necessary and sufficient causes. A weak sign goes from effect to sufficient cause which, although it carries a degree of necessity, is still too broad and must be narrowed down. However, both diagnostic signs (which go back from the effect to the cause) and prognostic signs (which go form the cause to the possible effects) rely on equally weak causes. Therefore, the conditions of a necessity of a sign are socially determined, according to what codes a culture determines to be strong.
In [1.11], Eco looks at how to look at how cultures develop codes. This usually occurs through “daring inferences” especially through abduction, which is the tentative tracing of a system of signification rules which will allow a sign to acquire meaning. We look to a general frame because all induction (the process of deriving general principles form particular facts or instances) is governed by the coded rules of a broader system. In general there are three types of abduction:
First, there is hypothesis or an overcoded abduction which is when the law seems to be a given. This type relies on presuppositions that occur within the co-textual environment.
Second, there is an undercoded abduction where rules must be selected from a series of equiprobable alternatives.
Third, there is creative abduction in which a brand new explanation is developed (“invented ex nono“).
All of this leads to [1.12] and Eco’s criterion of interpretability which is that substitution (as in aliquid stat pro aliquo) is not only necessary for a sign, the possibility of interpretation is necessary as well. With this criterion we are able to move from the study of specific signs to a study of general semiotics. This is the case because interpreting a sign is to define the portion of the continuum which serves as its vehicle in its relationship with the other portions of the continuum. Thus, interpreting a sign is defining a portion through the use of other portions. Consequently, there is no death of the sign, only the death of signs that are based on equivalence.
Thus, Eco concludes in [1.13] with a comment on signs and subjects. Having rejecting signs premised upon equivalency (or equality/identity) he sees the sign as the locus for the semiosic process, and the instrument through which the subject is endlessly made and remade. And this is important because the map of semiosis, as defined by a given historical stage, tells us who we are and what to think.
Well, that was a mouthful. I doubt anybody will read this.

Well, between reading George Lindbeck and completing a course on sacramental theology, I’ve been doing more and more reading about linguistics and semiotics (the theory and study of signs and symbols, especially as elements of language and other systems of communication). I am currently working on a paper called: “Theology as Presencing: Speaking Religion with George Lindbeck, Martin Heidegger, and Umberto Eco” and as a part of my research I’m reading Umberto Eco’s book, Semiotics and the Philosophy of Language. His work is decidedly dense — particular for those who don’t “speak the language” of language scholars. So, to help me work through this book, I’ve decided to start a series of posts summarising and reflecting on each chapter. That might not be terribly exciting for those who choose to read this blog but I figure if I can present Eco’s material in a way that makes sense to people not interested in this topic then I’m probably getting what he is talking about. So, if you’re interested, let me know what you think (Rivers? I’m counting on you here!). And if you’re not, well, too bad (that probably means everybody but Rivers). Without further ado:
[0] Introduction
Eco begins his book by stating that there tends to be two objects that are focused upon as the object of a general semiotic approach, and these two objects are generally seen as incompatible. They are:
1. The sign/sign-fuction which is a correlation between a signifier and a signified (i.e. between expression and content) and thus an action between pairs.
2. Semiosis which is not an action between pairs, but an action between three subjcts: a sign, its object, and its interpretant.
Eco’s thesis is that these two approaches are not incompatible. In order to defend this he will argue that the semiosic process of interpretation (emphasised by point 2) is actually also at the core of the concept of “sign” (point 1). Eco attacks the notion of signs as immobile or static notions, but argues that our fundamental approach to signs always involves an act of interpretation.
Having laid out his thesis, Eco then lays out some major themes that will be important throughout the book — and which will become more clear when I get to the chapters in which he addresses the individual themes.
Beginning with a basic principle of interpretation that says, “a sign is something by knowing which we know something more” (he is quoting Peirce at this point) and this implies an infinite process of interpretation. However, Eco argues that, although this process may be infinite in theory, any given text does not have an infinite amount of meanings. This is so because, when approached in light of a given topic (i.e. contextually) there is on a limited number of possible meanings. So Eco rejects the notion that there is not true meaning within a text (or, as Valery puts it, “il n’y a pas de vrai sens d’un texts“).
This means that Eco sees contemporary theories of interpretation on a line between two extremes, x and y.
At x are those who see only one possible way of interpreting the text — according to the author’s original intention.
At y are those who see any possible meaning in the text.
Eco is interested in finding a continuum of intermediate positions between these two points. Thus, his focus on context leads him to argue that between x and y stands “a recorded thesaurus of encyclopedic competence, a social storage of world knowledge” and interpretation is implemented and legitimated by this. The notion of interpretation being an act that is accomplished through an appeal to an encyclopedia of world knowledge is crucial to Eco’s thesis, as will become clear in our summary of chapter two.
Eco then goes on to cover some more introductory type material before getting into the body of his book. First, he explains that he is engaged in a philosophy of language because any general semiotics is a philosophy of language, and good philosophies of language are concerned with semiotics (see the definition of “semiotics” I provided above). Second, he makes a distinction between a specific semiotics and a general semiotics.
A specific semiotics is “the grammar of a particular sign system”. Thus, if one views Christianity as a particular kind of language, Christian theology functions as the grammar of that language — i.e. as a specific semiotics (I’m pulling on George Lindbeck’s work in this example). Thus, a specific semiotic is successful if it describes a given field of communicative phenomena as ruled by a specific system of signification. That is to say, a specific semiotic is successful if it forms a coherent worldview. That is why specific semiotics are used to improve, preserve, or destroy cultures. It is also why they are disciplinary, and they can tell which expressions are acceptable (i.e. grammatical) within that worldview. Thus, something can be empirically tested within a specific semiotics. In this regard, it can be said that specific semiotics is a science, because it has clear rules for how it determines the relevance of empirical data.
A general semiotics asks bigger questions: what does it mean to say, to express meanings, to convey ideas, or to mention states of the world? However, general semiotics will always be comparative, and rely on specific semiotics, since they are inescapable because all of us are contextual interpreters of meaning. We can only study and describe language through language. A general semiotics is not a science, but rather a philosophy because it cannot be empirically tested. This is so because philosophical entities only exist to the extent that they have been philosophically posited — that is to say, they are not ’emic’ (internal) definitions of previously recognisable ‘etic’ (external) data. Such concepts only have unity within their philosophical framework. Thus, a general semiotics has explanatory and practical power but cannot be shown to be true in a scientific sense. Yet this is not to say that such concepts are just figments of our imagination, but they cannot be judged for their truth-value, only for their perspicacity.
Okay… that’s it for the introduction. How’d I do? Clear as mud? Are you sitting there wondering why the hell anybody would want to care about this sort of stuff? Should I write more about why I think it’s interesting/important?