We must admit the existence of psychological
doubles of formal grammatical categories.
Lev Vygotsky1
One of the most impressive of Yeats's early poems is Who Goes with Fergus? By its intensity and power over language, it stands out from its flocculent neighbours; James Joyce has it play a prominent role in Ulysses; one of my teachers, William York Tindall, insisted that all of his students memorize it.
Who will go drive with Fergus now,
And pierce the deep wood's woven shade,
And dance upon the level shore?
Young man, lift up your russet brow,
And lift your tender eyelids, maid,
And brood on hopes and fear no more.
And no more turn aside and brood
Upon love's bitter mystery;
For Fergus rules the brazen cars,
And rules the shadows of the wood,
And the white breast of the dim sea,
And all dishevelled wandering stars.
Assuming that the poem deserves special notice among the myriad verse passages of western civilization, what is a critic to say of it? Immediately the great critics and scholars begin to loom from the shades. They proffer information: about the author, about his milieu, about Fergus, about metre and rhyme, about the (presumed) reaction of the (ideal) audience. Finally, the ‘formalist’ critics proffer information about ‘the work itself’, which in practice aims at a close description of the way the language of the poem works. If formalist analyses focus upon language, the latest formalists, the linguistic critics, seem to offer tools for the most detailed description of it.
An analysis by a linguistic critic would then proceed explicitly or implicitly on the assumption that there are many effects discernible in the poem which can be traced to the reader's response to the various configurations of language in it. An additional assumption, based on the first, is that the particular excellence of the work derives not from its content, from what is said, but either from the abstract deployment of the language itself, or from the closeness of the ‘fit’ between the content and the linguistic expression of that content: on the way in which what is said is said. Any ‘linguistic’ analysis will therefore turn out to be an analysis of style, strictly separating what is said from how it is said, or, in the terms used above, separating identificative base from stylistic performance.
‘Style’ here has a number of meanings. It seems to refer to a subjective impression in the reader of the stamp of another personality on the language involved. It also suggests the total effect of the linguistic structures as such, whether they form a portrait of the artist or contribute to the general abstract web of language. Let us see what further, more detailed, comments a linguistic critic might want to make along these lines.
He would note a certain oddness in the language of Fergus, as compared to that used in ‘ordinary’ circumstances.2 For instance, it contains a great many repetitions of sound and stress, many more than would occur in casual speech. This, when noted by the reader, identifies it as a poem. However, there are other oddities about it. At first glance there seems to be an excess of conjunctions. What is the function of all those ‘ands’? And then, there are a number of ordinarily ‘referential’ elements which are here free-floating: definite articles with no reference to previous descriptions, addresses to unintroduced personalities, references to un-described entities. What deep wood? What level shore? Who is Fergus? What young man? What maid? What brazen cars? The speaker of the poem seems to be quite willing to mystify his audience in such matters.
A reader of Yeats's early poetry would be able to say that such vagueness of reference is not simply the characteristic of the speaker of Fergus but of most of Yeats's early work. It might perhaps be characteristic of the young Yeats himself. The point might be made that such lack of respect for ordinary principles of topic-definition is a characteristic of certain nineteenth-century schools of poetry – Pre-Raphaelites, Symbolists, Decadents – and conveys a feeling of internality, a notion that the reader is always overhearing an internal discourse of the poet.
But what is actually happening? To an ordinarily sensitive reader the focus of the poem seems diffuse. The putative ‘frame of mind’ of the speaker is only negatively described by saying that the speaker seems willing to mystify the reader. The poem consists of questions, commands, and strong assertions. Why then is the effect so mild and Pre-Raphaelite? Why does the speaker's mind seem to be on something other than what he is saying? And why have the questions and imperatives in the poem the mitigated authority of the gestures of Beardsley figures?
A linguistic critic might begin his analysis of these matters by asking: what is a question, or a command, or an assertion? There has been a great deal of recent work on the conditions that define the asking of a true question, the issuing of a true command, the declaration of a truly intended assertion, following the lead of John Austin and John Searle and their description of the nature of ‘speech acts’ and ‘illocutionary force’.3 The ‘illocutionary force’ of an utterance lies in its intended use; the locution achieves its end only if its ‘constitutive conditions’ are satisfied. The constitutive conditions are different for the asking of ‘true’ questions, as opposed to rhetorical questions, examination questions, or riddles, to say nothing of locutions which have the form of questions without their force, such as ‘Why don't you go jump in the lake?’
Searle describes the conditions for the asking of true questions as follows:
(a) The questioner does not have the information necessary to complete a proposition truly.
(b) It is not obvious to the questioner and his audience that the audience will provide the information without being asked.
(c) The questioner wants the information.
(d) The question counts as an attempt to elicit information.4
The complex question under examination – ‘Who will go drive with Fergus now / And pierce the deep wood's woven shade, / And dance upon the level shore?’ – ‘fails’ as a true question by the absence of an immediately present audience, which seems to violate conditions (b) and (d). All written questions, with the exception of those in personal letters and government forms, would fail as true questions for these reasons. (In fact, Richard Ohmann has suggested that it is a general characteristic of literature at large, that the locutions that comprise it have all lost their illocutionary force in order to acquire the force specifically of literary locutions.5)
With this evidence that the question in lines 1–3 of Fergus is not a ‘true’ question, it becomes necessary to ask what it is, and to see if the answer is valuable for a critical description of the poem's style.
There is, as I suggested above, a general stylistic effect of mild absent-mindedness in the poem, the sense of a sort of double or contrapuntal mental ‘set’. The denial of all illocutionary force to the question in lines 1–3 would not convey this complex impression, at least by itself; the (written) questions in Blake's Tyger, for instance, convey quite a different impression.6 The generalized loss of illocutionary force by written or literary questions is not a genuine loss. It is as if the poet and the reader are playing a literary game in which the ‘penalty’ for violating the laws of asking true questions is the establishment of a limited situation. In that situation the process of asking and answering questions can only be imitated. The same rules apply within the imitative game, however, as in the non-literary context. The result of the game is not a true answering of questions but an observing of the asking of them, and the same would be true of imperatives and assertions: within the literary game, the reader observes the poet commanding and asserting, without feeling bound to obey or believe. However, since within the game the rules for asking questions, making commands or asserting are still being ‘obeyed’, we can use the linguist's tools to examine Fergus, as if we were analysing casual speech; we will always return at the end, however, to the moderated form of the language established by the literary game.
Illocutionary force can usually be determined clearly only by consultation with context, both linguistic and extra-linguistic. However, there are certain syntactic elements which inherently can have no place in either questions or commands, if these are to remain ‘purely’ questioning or commanding. These are the so-called ‘non-restrictive modifiers’, of which non-restrictive adjectives and relative clauses are the most commonly described type. In Othello's command ‘keep up your bright swords or the dew will rust them’ (I. ii. lines 58–9), the adjective ‘bright’ is non-restrictive, a type of parenthetical observation, with a definite ‘by-the-way’ feeling about it. (The context must make it clear, of course, that the ‘bright swords’ are not being contrasted, restrictively, with dull ones.) Another of Othello's commands, ‘Silence that dreadful bell!’ (II. iii. line 166) contains a non-restrictive modifier, ‘dreadful’, and is therefore not a ‘pure’ command. ‘Dreadful’ adds an effective note of irritation, but is not strictly necessary for the command; there is only one bell to silence, and thus no need to define it further. Mitigated commands are common in casual speech, so their use in poetry is not unprecedented. However for a full description of their effect we need to examine both their nature and their function.
When the speaker in Fergus observes the depth of the wood, the weaving of the shade, the levelness of the shore, and so on, he appears to be engaged in mental acts which take the linguistic form of non-restrictive acts of attribution. However, the surface evidence of these acts – ‘deep’, ‘woven’, ‘level’, ‘russet’, ‘tender’, ‘bitter’, ‘brazen’, ‘white’, ‘dim’, ‘dishevelled’ (and perhaps ‘wandering’, which may, however, be restrictive) – are all embedded in sentences which serve other illocutionary functions than that of simple attribution, in that they are questions and commands. The mental stance for a ‘pure’ command, as for a true question, does not possess the self-informative quality of the act of attribution. Questions confess ignorance and a drive for externally supplied information: they imply a future, when puzzles will be solved. Commands also contain (implicit) references to a future state, a state more satisfactory to the commander than the present. Attributives, however, refer to the past, in which the evidence for the attribution has been gathered. These ten or eleven non-restrictive attributes, therefore, modify and mitigate the questions in the poem (within the game), by dragging down the future force of questions and commands.
Question |
Assertions |
Who will go drive with Fergus now, and pierce the wood's shade? |
the woods are deep the shade looks as if someone had woven it. |
This introduction of non-restrictive ‘impurities’ is a primary source for the feeling of diffuseness and lack of definite focus in Fergus. The linguist is able here to point to the means by which the speaker of the poem seems to be lingering in the present and past whilst at the same time striving to affect the future. The poetic game is here played with the asking of questions, issuing of commands, and making of assertions – all of which are mitigated, in themselves and in context.
The use of linguistic techniques to provide criticism with additional tools of analysis is one type of description of style. Here the tools have helped us analyse Yeats's style, though at this point no distinction has been made between the analysis of the ‘poetic’ use of general language, in a game, and the description of the intensely personal element in Yeats's use of language. As we will see, this second ‘personal’ sort of analysis is based firmly on the first, for to be able to describe particular patterns of linguistic habits in the work of individuals, a close description of these habits must be available.
In describing the relationship between thought and language, the Russian psycholinguist Lev Vygotsky described language as the culmination of a series of internal psychological processes.
The word forms the end and not the beginning of the development. The word is the end which crowns the deed.7
In another place, Vygotsky compared thought to ‘a cloud shedding a shower of words’.8 Both of these views find in language the most intimate possible revelation of prior internal mental states. An examination of linguistic ‘output’ should, therefore, reveal some imprint of those deeply personal aspects of style which preceded and moulded it.
Richard Ohmann, in a well-known essay,9 has put forward a hypothesis that personal literary style can be described, at least in part, by examination of the characteristic idiosyncratic choice of syntactic structures employed by the author. For example, he finds in D. H. Lawrence constant and characteristic features of repetition and reduction; in the following passage from Studies in Classic American Literature, Ohmann discerns the obsessive personality of Lawrence emerging as an aspect of his style:
The renegade hates life itself. He wants the death of life. So do these many ‘reformers’ and ‘idealists’ who glorify the savages in America. They are death-birds, life-haters. Renegades.
We can't go back. And Melville couldn't. Much as he hated the civilized humanity he knew. He couldn't go back to the savages. He wanted to. He tried to. And he couldn't. Because in the first place it made him sick.
When Ohmann examines this passage, which features a high degree of elision, he finds it possible to reconstruct some of the syntactic elements which by their combination produce the passage, as well as the abbreviative conventions that disguise the extreme obsessive repetitiveness of the whole. With the restoration of the elided elements, the passage reads:
The renegade hates life itself. He wants the death of life. So do these many ‘reformers’ and ‘idealist’ who glorify the savages in America [want the death of life]. They are death-birds. [They are] life-haters. [They are] renegades.
We can't go back. And Melville couldn't [go back]. [Melville couldn't go back, as] much as he hated the civilized humanity he knew. He couldn't go back to the savages. He wanted to [go back to the savages]. He tried to [go back to the savages]. And he couldn't [go back to the savages].
Another aspect of this type of analysis deals with the characteristic choice of words rather than syntactic units. In the passage from Lawrence, there seems to be a deliberate attempt to choose words for their shock value, in the place of their more decorous and abstract synonyms. The passage would read very differently if for ‘savages’ we substituted the currently fashionable phrase ‘non-technologically advanced peoples’. Use of abstraction or euphemism (or its opposite, dysphemism, for that matter) seems to be as much a personal stylistic trait as a social one.
There are other possible sorts of ‘personal’ stylistic characterization through language. For example, a typically ‘logical’ order to discourse by topic and subtopic, even in a casual context, can characterize a person to those of his hearers who are sensitive to such rhetorical clarity. He may also be distinguished by choice of topic, as well as the appropriateness (or the reverse) of the topic to the context chosen. In this case both the topic itself and the speaker's sense of congruity to social surroundings are factors in a judgement of style. In addition to these internal linguistic factors, there are those which depend upon external physical ‘givens’, the physical signs of individual style: quality of voice, appearance and deployment of physical appearance and ‘body-set’, idiosyncratic penmanship, and the like. Every aspect of descriptive linguistics – graphemics, kinesics, paralanguage, phonology, syntax, semantics, rhetoric, sociolinguistics and psycholinguistics, among others – is capable of distinguishing style.
Literary style, like all style, reveals both the individual personality of the creator and the general language habits of the society in which he lives. To see how such a double focus is possible, we should examine our ways of interpreting the evidence of our senses – our public and private strategies of perception.