Reading Waugh on Jakobson

Waugh, Linda R. 1980. The Poetic Function in the Theory of Roman Jakobson. Poetics Today 2(1): 57-82.

As Roman Jakobson has shown (1960), if we take any given act of verbal communication (= speech event), there are six fundamental factors which must be present for it to be operatble:

1. addresser (speaker, encoder, emitter; poet, author; narrator)
2. addressee (decoder, hearer, listener; reader; interpreter)
3. code (system, langue)
4. message (semelfactive parole, the given discourse, the text)
5. context (referent)
6. contact ("a physical channel and psychological connection between the speaker and addressee").

Each of these factors may be further subdivided and separated out in various ways. For example, in a literary text, there may be a differentiation between author and narrator and the narrator in turn may give way to several other 'speakers.' In addition, the lack of or throuble with or ambiguity about any one of the factors may have various effects on the communication itself. For example, it should be noted that without a message, there is no act of verbal communication. (Waugh 1980: 57-58)

That is to say, the verbal communication act is not only verbal, but vocal. Weirdly, the tone or paralinguistic quality of the vocal message is not within the message, but on the side of the addresser, in the emotive function. The last bit, that there is no act of verbal communication without a message seems odd, considering the communicative value of silence (a topic on which many books have been written. And "semelfactive" is a lexical aspect in linguistics. Semelfactive verbs are punctual, instantaneous, momentive, like "blink", "sneeze" and "knock".

Corresponding to these six factors are six major functions, each assuming an orientation within the verbal message on one of the factors:

1. emotive (expressive)
2. conative (appelative)
3. metalingual (metalinguistic, 'glossing')
4. poetic (aesthetic)
5. referential (cognitive, denotative, ideational)
6. phatic

(For further discussion and elaboration, see RJ, 1960 and Holenstein, 1979 [in Semiotica]. In this formulation, the poetic function comprises the focus within the verbal message on the verbal message itself. It is, in fact, "un langage qui met l'accent sur le langage" (Todorov, 1978). (Waugh 1980: 58)

And indeed Waugh identifies the poetic function with the aesthetic function (e.g. Jan Mukarovsky 1936. Aesthetic Function, Norm and Value as Social Facts). Tzvetan Todorov's french utterance translates as "language that focuses on language". It is beginning to seem that most functions in Jakobsons model are merely the metalingual function in disguise.

As evidence of the unmarked nature of the referential function, we may cite the fact that in many linguistic and philosophical studies of language, the referential function has been said to be the only function of language; or, if (some of) the other functions have been discerned, they have been declared to be 'deviant' or 'unusual' or needing special consideration. And even in our parlance about language, the referential function is spoken of as 'ordinary language.' (Waugh 1980: 58)

This may be because early study of language was primarily the playground of the philosophy of language, which concentrated on the reference because it is somewhat more concrete than other functions. Phatic function came from social anthropology, the emotive function belongs to social psychology, etc. In this regard Jakobson did manage to bring semantics into play with aspects that were studied by non-semanticists.

And in fact, one could make a typology based on which other function(s) assume some importance, as in the following statement by Jakobson: "Epic poetry, focussed on the third person, strongly involves the referential function of language; the lyric, oriented toward the first person, is intimately linked with the emotive function; poetry of the second person is imbued with the conative function and is either supplicatory of exhortative, depending on whether the first person is subordinated to the second or the second to the first" (RJ, 1960:357). (Waugh 1980: 59)

This sounds like the type of general typologization that Juri Lotman occupied himself with (e.g. cultures that orient towards mnemonic autocommunication versus those that orient towards non-mnemonic, etc.).

1.20 The set toward the message brings forth the question of the nature of the message per se.
1.21 In answer to a statement by Degas that he was full of ideas but couldn't manage to say what he wanted to say in a poem, Mallarme's reply was: "My dear Degas, one does not make poetry with ideas but with words" (Valery, 1939/1958:63). We may modernize his reply and state that poetry is made not with ideas but with signs, words being only one of the types of signs (and further state that poetry is not about the real world or life, but about itself). (Waugh 1980: 60)

What is there to question? It is a model of verbal communication act. The nature of the message is much more interesting in Ruesch, who says that most anything can serve as a message - be it an object, action or word - much like for Peirce anything whatsoever can act as a sign in some degree or measure for some interpreter (I'm paraphrasing). Questioning whether the verbal message is a word, idea or sign is trivial, because it is linguocentric. It would suffice to say generally: it is verbal.

Language - both code and message - is a system of systems of signs, a sign being an intrinsic and indissoluble combination of a perceptible signans and an interpretable signatum. Some linguistic signs occur both in the code and in messages; others occur only in messages. In fact, given this definition of sign, the message itself is a system of systems of signs and at the same time a sign (of some complexity) with both a signans and a signatum. The act of verbal communication is, in effect, an exchange of signs between speaker and addressee. (Waugh 1980: 60)

I would argue that the code, being constructed or invented and finite in its scope, is a system of signs, and language, which is composed of a variety of subcodes, dialects, idiolects, etc., is a system of systems of signs. But then, if you consider that culture is a composite of languages, codes and texts, you would have to say that culture is a system of systems of systems of signs. You could probably add another "system of" if you consider the semiosphere to be the totality of cultures. All in all, this quote also verifies that in some cases texts are viewed as systems of signs, and obviously it is a justified perspective - especially in relation with other systems (other texts) and other "extra-textual" connections, indeed any texts whatsoever becomes a system. But then again, it may be the case that I'm stretching the meaning of "system" here.

While, in particular, sentences, utterances, and discourses (i.e., those higher in the hierarchy), may be unique signs which have never occurred before, they are nevertheless intersubjective signs common to speaker and addressee. They may be transitory signs, especially if they occur in spoken language, and semelfactive signs, reoccurring at various times but without necessarily being codified; but they are nevertheless signs, capable of being interpreted, due to the code which guarantees that interpretation. (Waugh 1980: 61)

When I began reading semiotics I was frustrated by the amount of "semiotic whatevers" like semiotic routine, value, unit, phenomena, intensity, space, capacity, reality, etc. Almost anything seemed to become semiotic with the simple addition of "semiotic" in front of it. Here we have a case of lexical aspects becoming types of signs. While other categories are to some degree available in Morris's semiotic terminology (e.g. interpersonal signs), I find "semelfactive signs" just annoying. Much like Ecos "inchoative semes", it uses a linguistic term metaphorically for theorizing signs. In the original definition, semelfactive signs should signify actions that are punctual-instantaneous-momentary; here they are merely signs that occur punctually-instantaneously-momentarily. That is, Waugh jumped from the semantic dimension onto the syntactic one. If we wished, we could add a pragmatic dimension: semelfactive signs are interpreted punctually-instantaneously-momentarily (e.g. something catches your eye and ear and you take momentary notice but something interrupts further thought about it).

Now, if we take the hierarchy of signs such as phrases, clauses, sentences, utterances, discourses, not only is this a part-whole hierarchy of ascending complexity, but also one of ascending freedom or creativity. While the sentence is the largest sign for which the rules of combination are obligatorily codified, it is still the case that the speaker has more freedom in creating sentences than in creating phrases. Of course, there are more sentence types than phrase types (i.e., more different syntactic matrices of sentences than of phrases) and the number of possible sentences is a function of the number of possible phrases and as well of the number of possible combinations of those phrases. Consequently, the freedom of the speaker to be 'creative,' to construct new and unique messages (i.e., new signs) is greater. (Waugh 1980: 62)

I think there may be some little heuristic value in looking at the varieties of signs on the internet. In #netisemiootika, these would not be organized on the principle of hierarchy and ascending complexity, but on something else that I don't know how to name at the moment. Starting with abbreviation, short interjections and emoticons (lol, wat, :D), then moving on to Waugh''s series: words, phrases, clauses, sentences, utterances, etc. and then proceeding to links and embedded objects: pictures, video players, sound players, and links to an endless variety of mixed-media websites. Multimedia has enabled a level of creative freedom unforeseen: drawings seen by both parties in real time, snapchats, video calls, etc. On the other hand this precept on freedom could very well be applied to web interactions: comment forms have character limits (tweets cannot exceed 140 characters). Some tumblrs ever write their witty remarks at length purposefully so that their remarks cannot be simply copy-pasted on twitter.

Since a poem is a sign, it is a combination of a signans and a signatum; it is not, as some studies of poetry have seemed to suggest, a signans only. A poem is a new intersubective message-sign whose dominant function is an orientation toward the message-sign as a message-sign. A poem is also a system of systems of signs, a complex and hierarchically ordered sign, made up of a variety of sign types, each with both a signans and a signatum, and in which the various signs are subordinated to the overall poetic function and coherence of the whole sign. As such, a poem is a structure, "not a mechanical agglomeration but a structural whole and the basic task is to reveal the inner [...] laws of this system" (RJ, 1972:711). These words, while they were written in 1929 by Jakobson about structuralism fully define, in general, the poem. (Waugh 1980: 62)

Yeah, but all of this goes for almost every type of text - complexity and variety of sign types is inevitable (one could even say, as Juri Lotman perhaps even did, that a simple repetition of the same sign is not a text). An orientation toward the message-sign as a message-sign is above hinted at by stating that a poem 'announces' that it is a poem. This is similar to Mukarovsky's treatment of the "artistic object" which necessitates an interpreter who is able to interpretate a physical fact (a book, a painting) as an aesthetic object. But once again, this is the function of almost every type of sign: a text announces that it is a text, a smile announces that it is a smile, etc. One way to approach it would be through S. K. Langers distinction between signification and significance (which she probably borrowed from Lady Welbys significs): a text has significance, but the poetic function announces that it is a specific type of texts and has significance as a poem. If applied to communication and messages in general, then this significance comes close to metacommunicative cues: the message contains implicit instructions that announce that it is indeed a message and indicate how the content of the message should be interpreted. But alas MC cues are more likely involved in what Welby places as an intermediary between "sense or signification" and "significance or ideal worth": namely "meaning or intention".

The 'function' of a given message is, in Jakobson's terminology, an intrinsic quality of the message itself; this, the focus upon the message is an inherent quality of a poem. (Waugh 1980: 62)

I just thought that maybe this function is present in nonverbal communication in the "directedness" of certain messages. Namely, looking someone in the eye and performing a certain facial or hand gesture at the same time directs attention to said gesture. It says as if "look at this". A common example would be of someone mimicking someone's facial expression and conveying this to a companion. As if "s/he did this face".

The verbal material displays overall a hierarchical structure of symmetries, based on repetitions, regularities, and systematizations of various kinds. There is, in other words, a radical parallelistic reorientation of all the verbal material as it relates to the building of the sequence. Such parallelisms, whether based on sound (see RJ, 1979) or on grammatical categories (see RJ, 1966b) or on various lexical categories, are a 'natural' result of the raising of equivalence to the constitutive device of the sequence, as against any sort of non-poetic counterparts (see Lodge, 1977). Moreover, such parallelisms create a network of internal relations within the poem itself, making the poem into an integrated whole and underlining the poem's relative autonomy. (Waugh 1980: 64)

Something similar occurs in most texts. Especially scientific texts (and particularly theoretical and social science papers) regularly use the device of repetition. For example, in the theoretical portion of the paper some notions are defined as important and in the empirical/analytical portion these notions are regularly repeated in a certain order. E.g. "we must consider here the emotive, conative, metalingual, referential and phatic functions in relation with the poetic function." Usually, though, there are three items that are repeated (icon, index, symbol; Umwelt, Innenwelt, Lebenswelt; individual, socio-cultural, biological; etc.).

In referential speech, the relation between the linguistic sign (in particular, the signatum) and extra-linguistic 'objects' (sometimes called referents) would seem to be a close and almost automatic one. In the referential use of language, the word is evaluated as a proxy for the denoted object, or for the idea/concept and in emotive use as an outburst of an emotion - the word is, so to speak, a 'verbal shadow' of the object or of an idea/concept or of an ambition (see Erlich, 1955:187). Of course, we should not let the terminology 'object,' 'idea,' 'emotion' lead us astray; as C. S. Peirce has shown, emotions and ideas are themselves signs (albeit potentially non-linguistic signs - see Savan, 1980), while objects are generally semiotic or capable of being semioticized (see Waugh, 1979c). The bond between word and object or word and emotion then is the bond between a linguistic sign and another sign (from another semiotic system). (Of course, in referential language, the bond is one which was called 'arbitrary' by Saussure, while in emotive language it is based on intimacy - see 3.2 - and in particular on a similarity relation - see RJ, 1980b; Waugh, 1979c.) One may say, with Peirce, that in the referential or emotive use of language, the linguistic sign and the non-linguistic sign are in interpretive relation with one another. It is this status of interpretation which is broken in poetic discourse, for here it is the interpretive relation between linguistic signs which is important. (Waugh 1980: 67)

We must keep in mind that the "object" in Peirce's triadic sign model is itself another sign. "Extra-linguistic" doesn't make sense from that perspective.

Thus, a 'literal' (i.e., a referential) reading of a poem must be subordinated to a 'poetic' reading (just as the referential function is subordinated to the poetic). (Waugh 1980: 68)

Here it would be appropriate to substitute poetic with aesthetic: the artistic text doesn't refer to extra-textual reality, but functions as an aesthetic object for appreciation.

Sound as such, in and of itself, becomes one of the patent carriers of poetic meaning: there is a kind of 'verbal magic' is sound itself. Various sound figures are used for the construction of the sequence. (Waugh 1980: 69)

I think I'll reserve "verbal magic" for occurrences in which language does something incredible (e.g. the case of verbal influence). Or, alternatively, for Hall's 'language band', which is very general but still relevant (when someone is speaking it is rude to stop listening and just walk away, and some people abuse this simplistic type of "magic" and persuade others simply be speaking in an uninterruptible manner).

The actualization of the linguistic sign as a sign leads also to a focus upon the signatum both in its relation to the signans and in terms of its own structural nature. If we take the signatum of any linguistic sign, it evidences a continual dichotomy between a relational invariant - sometimes called a general meaning (Gesamtbedeutung) - and a variety of hierarchized contextual variants (specific meanings) including the basic variant (= Grundbedeuting), that is, that variant which is 'typical' for the given sign and is least conditioned by the environment. (Waugh 1980: 71)

My first thought was that Morris hasn't anything like this. But then I remembered that he did distinguish general signs from singular signs. Signs also vary in their generality. But this is not exactly what is presented here. Rather, it reminds me of Barthes's interpretation of denotation and connotation.

Now, while the potential for ambiguity resides in the relation of contextual variants to the invariant, generally contextual meanings are discernible both from surrounding linguistic context (e.g., other words in the same discourse) and from the general situation (the non-verbalized context but capable of being verbalized) in which the speech event takes place. (Waugh 1980: 72)

Why.

A poem is to a certain degree decontextualized: it is a system of systems which is more self-contained than referential discourse. One could say that the poem provides its own 'universe of discourse.' The orientation of the poem upon itself as a message-sign has the effect of making more of a break between the poem and its context than in referential discourse and results in a relative self-suffuciency for the poetic text. (Waugh 1980: 72)

It seems to me that this is also the case with dystopian literature, which is oriented to the future and thus quite separated from the events of real history.

But a poem does not exist in a vacuum: it is part of a general historico-cultural context and indeed depends on that context for its interpretation. Nor is it sealed off from a literary context. (Waugh 1980: 72)

This is also true for dystopian literature. Although separated from real history, every dystopic work has something in common with other dystopic worlds - e.g. the case of genre memory.

Redundant signs are those signs which inform about other signs in the text and thus cannot be said to provide independent information; they are used in a sense to ensure that the given information is provided. (Waugh 1980: 73)

I think Charles Morris calls these metasigns.

Ambiguous signs are those which, even when in context of other signs, provide more than one interpretation; the potential for such ambiguity is inherent to any general meaning. (Waugh 1980: 73)

Morris similarly talks about ambiguous and unambiguous sign-vehicles.

Waugh, Linda R. 1998. Marked and unmarked: A choice between unequals in semiotic structure. Semiotica 38(3-4): 299-318.

Apparently any (or might it not be 'any'?) phonological correlation acquires in the linguistic consciousness the form of a contraposition of the presence of a certain mark to it absence (or of the maximum of a certain mark to its minimum). Thus, one of the terms of the correlative necessarily proves to be 'positive', 'active', and the other becomes 'negative', 'passive'. (Trubetzkoy 1975: 162f in Waugh 1998: 300)

Oh wow. This is the clearest formulation on this topic that I have come across. In nonverbal study this kind of markedness theory could possibly applied profitably in relation with presence/avoidance and attraction/aversion. Simply put: instead of looking only at what is there also consider what is not there.

I am coming increasingly to the conviction that your thought about correlation as a constant mutual connection between a marked and unmarked type is one of your most remarkable and fruitful ideas. It seems to me that it has a significance not only for linguistics but also for ethnology and the history of culture, and that such historico-cultural correlations as the life〜death, liberty〜non-liberty, sin〜virtue, holidays〜working days, etc., are always confined to relations a〜non-a, and that it is important to find out for any epoch, group, nation, etc., what the marked element is. (Trubetzkoy 1975: 162f in Waugh 1998: 300)

It seems that Juri Lotman tried to realize the potential of this idea for the study of the history of culture.

...the tendency in implicational statements for the unmarked term to be implicatum and the marked term to be the implicans; (Waugh 1998: 302)

I did not know that the concept of implication can be used this way, e.g. differentiating the signans and signatum sides of the implication. Maybe this idea is so stunning because commonsense conception of implication says that it doesn't have its autonomous signifier, but that it is a kind of co-signatum.

This metaphorical rendition helps to point up the fact that there are two structurally given interpretations of the unmarked term - interpretations I will characterize as the 'zero' interpretation (the set/ground) and the 'minus' interpretation (the set minus the subset or the ground minus the figure) (See Figure 2).

The 'zero-interpretation' (called by Jakobson [1932, 1936, 1939] the 'zero-sign' and also the 'non-signalization of x') is that interpretation that is the most general, widest, and most broad; that interpretation wheret he presence or absence of the unit of information is for the most part irrelevant. (Waugh 1998: 302-303)

Danesi explicated these versions in his paper on binary oppositions as well (in SSS). I feel as though Waughs models could be compared to Juri Lotman's semiosphere model and its development, to make sense of the semiotic, other-semiotic and non-semiotic. Also, I would add that "'minus' interpretation" is not the best term one could come up with. For sake of convenience, I would term it minus-marked (miinusmärgistatud) instead of unmarked. Also, I'm not at all sure where this theory came from, but it may be that Trubetzkoy read Zamjatin's We (as it was not published in Russian but it was published in Czech (?)). Namely, Zamjatin used his knowledge of mathematics very creatively and in his introduction to We says that his (dystopic) socio-fantasy is "preceded not by a plus but by a minus sign".

If we look at the much studied difference between "spoket" and "written" language, it would seem that in the context of the history of humanity as a whole, "spoken" language is the unmarked term and "written" language the marked term. "Written" language is more specialized in many ways than "spoken" language. We can see that quite clearly through the distribution of written and spoken language over different cultures: spoen language is universal, written language nonuniversal; there are many cultures with no written language, while all human communities have a spoken language. (Waugh 1998: 308)

I like writing more than speaking. You can write words that you can't pronounce. In a single culture, written language may be the same for all people while spoken language is very different from area to area (China, Ancient Egypt). Also, written word is recorded word, it lasts longer.

The "wearing of clothes" on social occasions is unmarked, whereas the "wearing of clothes" in a nudist colony or in the shower is marked. "Speaking in a normal tone of voice" is unmarked except in such special circumstances as in a library, in a church, etc., where it becomes marked. (Waugh 1998: 310)

This concerns the context dimension of metacommunication, e.g. situation baseline or definition of the situation.

In a culture like ours, in certain contexts, "life" in unmarked and "death" marked; and very often "life" is unmarked in the sense of the zero-interpretation. Generally speaking, we do not question at every moment of our lives whether the next moment should include life or death, or whether especially we should motivate our choice to live. Rather the assumption is that we will choose "life", or better (zero-interpretation), that "life" will simply contunie - there is no choice to be made. On the other hand, the choice of death - e.g., suicide - is the marked one from the point of view of the culture as a whole. Generally speaking, the choice of life or the continuing to live does not need a motivation, while the choice of death must have a motivation. (Waugh 1998: 313)

A good point. Suicide letters try to explicate those motivations. In the notes Waugh adds that Jakobson pointed out the reversion of this markedness relation in Majakovski: "for him life needed motivation and death didn't. Majakovski of course committed suicide." (ibid, 316)

Waugh, Linda R. 1997. Roman Jakobson's work as a dialogue: The dialogue as the basis of language, the dialogue as the basis of scientific work. Acta Linguistica Hafniensia: International Journal of Linguistics 29(1): 101-120.

While dialogue is an important facet of contemporary work in a number of fields, inspired especially by the work of Mikhail Bakhtin, few are aware of this facet of Roman Jakobson's work. (Waugh 1997: 101)

Jakobson is indeed not the first one that comes to mind in relation with the notion of dialogue. Even Juri Lotman seems to have more to do with dialogue, seeing as even his understanding of the concept of text is dialogical (we do not read the text but communicate or conversate with it).

In a famous formulation (1960b), Jakobson contended that linguistic communication rests on the speech event, which itself is composed of six facets: speaker, addressee, message, code, contact, context. (Waugh 1997: 101-102)

Waugh is perfectly aware that Jakobson's model is primarily a model of the "speech act", which is why she can freely replace "addresser" with "speaker". It is relevant though, that the addressee is not simply a "listener" but the person to whom speech is addressed, because the addressee in Jakobson's model is a passive (intended) destination of the message.

dialogue, for Jakobson, is thus the basis for language. And, in fact, in a series of lectures given in 1942, Jakobson spoke about the dialogue in a series of courses (in French) on Ferdinand de Saussure. The first part of those lectures dealt in a critical manner with the langue-parole distinction (which Jakobson later renamed code and message, after being inspired by communication theory - see Jakobson 1961). Although the lectures had beet written out when he gave them, he never published them; they were found after his death and published first in the original French in 1984 and then in English transration in 1990 as "Langue and Parole: Code and Message". The lectures argued for the basically dialoguc nature of language: for Jakobson, monologue is secondary and based on dialogue; and thinking as an interiorization is also based on the dialogue (with oneself). He argued that both language and parole (both the potential and the actual) are fundamentally social, because they are based on the dialogue, which is social by definition. They may also have a personal, individual side (personal styles and messages addressed to the self, for example), but this is an overlay, so to speak, on the social underpinnings. In other words, for Jakobson, the social nature of language takes precedence over the individual, just as dialogue takes precedence over monologue. (Waugh 1997: 102)

This is truly fascinating. Especially because of the year of its first publication, which coincided with Lotman's paper on the semiosphere (where dialogue is given a privileged position). The passage is also relevant for autocommunication.

For Jakobson, conversing not only dominates linguistic activity but also resides over scholarshipp Science is a dialogue, not a series of monologues (Jakobson 1933: 539), just as language is based fundamentally on the dialogue, not on a series of monologues. (Waugh 1997: 102)

With this I agree. It also dawned on me that my method of studying - quoting and commenting - occurs in the form of a passive dialogue. Almost every post in this blog is a small conversation.

...Jakobson rescued many important names from oblivion. This citation of so many other thinkers is the result not only of an enormous erudition (see, for example, Jakobson 1975a, on the medieval grammarians, and 1975b on the history of semiotics) but also of Jakobson's receptiveness to the ideas of his predecessors and openness to others in general. (Waugh 1997: 105)

Admirable virtues.