A method of analysis and some examples
This chapter has the practical purpose of showing how the apparatus of linguistic description can be
used in analysing the style of a prose text. We take for granted the principles that have been argued in
Chapters 1 and 2. We also take for granted a set of linguistic categories which will be more or less
common knowledge to those who have a basic familiarity with the workings of the English language,
whether in literary or non-literary contexts. One particular area in which technical terms are likely to
cause some problems is that of grammar, and here we follow the terminology and general view of
grammar presented in Greenbaum and Quirk’s A Student’s Grammar of the English Language and many
other grammars using a similar framework.1 Another area is that of foregrounding, where we draw on
the terminology of traditional poetics (‘metaphor’, ‘metonymy’, ‘onomatopoeia’, etc.). Although many
of these terms are widely current in literary scholarship, we presuppose a linguistic account of these
phenomena, and for this purpose, it is convenient to refer the reader to Leech’s A Linguistic Guide to
English Poetry.2 Following the list of categories in section 3.1, explanations of selected points will be
added in section 3.2.
Every analysis of style, in our terms, is an attempt to find the artistic principles underlying a writer’s
choice of language. All writers, and for that matter, all texts, have individual qualities. Therefore the
features which call themselves to our attention in one text will not necessarily be import-ant in another
text by the same or a different author. There is no infallible technique for selecting what is significant.
We have to make ourselves newly aware, for each text, of the artistic effect of the whole, and the way
lin-guistic details fit into this whole.
3.1 A checklist of linguistic and stylistic categories
The categories are placed under four general headings: lexical categories, grammatical categories,
figures of speech, and cohesion and context. Seman-tic categories are not listed separately, since, as
suggested in section 2.9, it is easier to arrive at these through other categories; for example, we use our
lexical categories to find out how choice of words involves various types of meaning. Since the purpose
of the list is heuristic, there is no harm in mixing categories in this way. It is also in the nature of things
that categories will overlap, so that the same feature may well be noted under different headings.
A: Lexical categories
1 GENERAL. Is the vocabulary simple or complex(i)? formal or colloquial? descriptive or
evaluative? general or specific? How far does the writer make use of the emotive and other associations
of words, as opposed to their referential meaning? Does the text contain idiomatic phrases or notable
collocations(ii), and if so, with what kind of dialect or register(iii) are these idioms or collocations
associated? Is there any use of rare or specialised vocabulary? Are any particular morphological
categories noteworthy (e.g. compound words, words with particular suffixes)? To what semantic fields
do words belong?
2 NOUNS. Are the nouns abstract or concrete? What kinds of abstract nouns occur (e.g. nouns
referring to events, perceptions, processes, moral qualities, social qualities)? What use is made of
proper names? Collective nouns?
3 ADJECTIVES. Are the adjectives frequent? To what kinds of attribute do adjectives refer?
Physical? Psychological? Visual? Auditory? Colour? Referential? Emotive? Evaluative? etc. Are adjectives
restrictive or non-restrictive? Gradable or non-gradable? Attributive or predicative?
4 VERBS. Do the verbs carry an important part of the meaning? Are they stative (referring to
states) or dynamic (referring to actions, events, etc.)? Do they ‘refer’ to movements, physical acts,
speech acts, psychological states or activities, perceptions, etc.? Are they transitive, intransitive, linking
(intensive), etc.? Are they factive or non-factive(iv)?
5 ADVERBS. Are adverbs frequent? What semantic functions do they perform (manner, place,
direction, time, degree, etc.)? Is there any significant use of sentence adverbs (conjuncts such as so,
therefore, however; disjuncts such as certainly, obviously, frankly)(v)?
B: Grammatical categories
1 SENTENCE TYPES. Does the author use only statements (declarative sentences), or do questions,
commands, exclamations or minor sentence types (such as sentences with no verb) also occur in the
text? If these other types appear, what is their function?
2 SENTENCE COMPLEXITY. Do sentences on the whole have a simple or a complex structure?
What is the average sentence length (in number of words)? What is the ratio of dependent to
independent clauses? Does complexity vary strikingly from one sentence to another? Is complexity
mainly due to (i) coordination, (ii) subordination, or (iii) parataxis (juxtaposition of clauses or other
equivalent structures)? In what parts of a sentence does complexity tend to occur? For instance, is there
any notable occurrence of anticipatory structure (e.g. of complex subjects preceding the verbs, of
dependent clauses preceding the subject of a main clause)(vi)?
3 CLAUSE TYPES. What types of dependent clause are favoured: relative clauses, adverbial clauses,
different types of nominal clauses (that-clauses, wh-clauses, etc.)? Are reduced or non-finite clauses
commonly used and, if so, of what type are they (infinitive clauses, -ing clauses, -ed clauses, verbless
clauses)(vii)?
4 CLAUSE STRUCTURE. Is there anything significant about clause elements (e.g. frequency of
objects, complements, adverbials; of transitive or intransitive verb constructions)(viii)? Are there any
unusual orderings (initial adverbials, fronting of object or complement, etc.)? Do special kinds of clause
construction occur (such as those with preparatory it or there)?
5 NOUN PHRASES. Are they relatively simple or complex? Where does the complexity lie (in
premodification by adjectives, nouns, etc., or in postmodification by prepositional phrases, relative
clauses, etc.)? Note occurrence of listings (e.g. sequences of adjectives), coordination or apposition.
6 VERB PHRASES. Are there any significant departures from the use of the simple past tense? For
example, notice occurrences and functions of the present tense; of the progressive aspect (e.g. was
lying); of the perfective
A method of analysis and some examples
aspect (e.g. has/had appeared); of modal auxiliaries (e.g. can, must, would, etc.). Look out for phrasal
verbs and how they are used.
7 OTHER PHRASE TYPES. Is there anything to be said about other phrase types: prepositional
phrases, adverb phrases, adjective phrases?
8 WORD CLASSES. Having already considered major or lexical word classes, we may here consider
minor word classes (‘function words’): prepositions, conjunctions, pronouns, determiners, auxiliaries,
interjections. Are par-ticular words of these types used for particular effect (e.g. the definite or
indefinite article; first person pronouns I, we, etc.; demonstratives such as this and that; negative words
such as not, nothing, no)(ix)?
9 GENERAL. Note here whether any general types of grammatical construc-tion are used to special
effect; e.g. comparative or superlative construc-tions; coordinative or listing constructions; parenthetical
constructions; appended or interpolated structures such as occur in casual speech. Do lists and
coordinations (e.g. lists of nouns) tend to occur with two, three or more than three members? Do the
coordinations, unlike the standard construction with one conjunction (sun, moon and stars), tend to
omit conjunctions (sun, moon, stars) or have more than one conjunction (sun and moon and stars)?
C: Figures of speech, etc.
Here we consider the incidence of features which are foregrounded (see section 1.4) by virtue of
departing in some way from general norms of communication by means of the language code; for
example, exploitation of regularities of formal patterning, or of deviations from the linguistic code. For
identifying such features, the traditional figures of speech (schemes and tropes) are often useful
categories.
1 GRAMMATICAL AND LEXICAL. Are there any cases of formal and structural repetition (anaphora,
parallelism, etc.) or of mirror-image patterns (chiasmus)? Is the rhetorical effect of these one of
antithesis, reinforce-ment, climax, anticlimax, etc.(x)?
2 PHONOLOGICAL SCHEMES. Are there any phonological patterns of rhyme, alliteration,
assonance, etc.? Are there any salient rhythmical patterns? Do vowel and consonant sounds pattern or
cluster in particular ways? How do these phonological features interact with meaning(xi)?
3 TROPES. Are there any obvious violations of, or departures from, the linguistic code? For
example, are there any neologisms (such as Americanly)? Deviant lexical collocations (such as
portentous infants)? Semantic, syntactic, phonological, or graphological deviations? Such deviations
(although they can occur in everyday speech and writing) will often be the clue to special interpretations
associated with tradi-tional poetic figures of speech such as metaphor, metonymy, synecdoche, paradox
and irony(xii). If such tropes occur, what kind of special inter-pretation is involved (e.g. metaphors can
be classified as personifying, animising, concretising, synaesthetic, etc.)? Because of its close connec-
tion with metaphor, simile may also be considered here. Does the text contain any similes, or similar
constructions (e.g. ‘as if’ constructions)? What dissimilar semantic fields are related through simile?
D: Context and cohesion
Finally, we take a preliminary look at features which will be more fully dealt with in Chapters 7 to 10.
Under COHESION, ways in which one part of a text is linked to another are considered: for example, the
ways in which sentences are connected. This is the internal organisation of the text. Under CONTEXT
(see the discussion of discourse situation in section 8.1) we consider the external relations of a text or a
part of a text, seeing it as a discourse presupposing a social relation between its participants (author and
reader; character and character, etc.), and a sharing by participants of knowledge and assumptions.
1 COHESION(xiii). Does the text contain logical or other links between sen-tences (e.g.
coordinating conjunctions, or linking adverbials)? Or does it tend to rely on implicit
connections of meaning? What sort of use is made of cross-reference by pronouns (she, it,
they, etc.)? by substitute forms (do, so, etc.), or ellipsis? Alternatively, is any use made of
elegant variation – the avoidance of repetition by the sub-stitution of a descriptive phrase
(as, for example, ‘the old lawyer’ or ‘her uncle’ may substitute for the repetition of an earlier
‘Mr Jones’)?
2 CONTEXT. Does the writer address the reader directly, or through the words or thoughts of
some fictional character? What linguistic clues (e.g. first-person pronouns I, me, my, mine) are there of
the addresser–addressee relationship? What attitude does the author imply towards his or her subject?
If a character’s words or thoughts are represented, is this done by direct quotation (direct speech), or by
some other method (e.g. indir-ect speech, free indirect speech)(xiv)? Are there significant changes of
style according to who is supposedly speaking or thinking the words on the page?
3.2 Notes on the categories
(i) In a formal sense, word complexity should be measured by counting morphemes. For example,
un-friend-li-ness contains four morphemes
A method of analysis and some examples
and war only one. But determining the number of morphemes in a word can be a problem, especially
with words of foreign or classical origin, such as signification. For this reason, counting the number of
syllables per word is a more convenient measure of complexity. Morphemic complexity and syllabic
complexity are in gross terms reasonably equivalent; but they are not necessarily equivalent for
individual words; for example, six-th-s contains three morphemes, but only one syllable; establish, on
the other hand, contains only one morpheme, but three syllables.
(ii) An idiom may be roughly defined as a sequence of two or more words, the meaning of which is
not predictable from the meanings of the constituent words; e.g. get by, as it were, under the weather.
A colloca-tion is a combination of words, which may be habitual (e.g. blue sea) or contrary to
expectation (e.g. mad sea).
(iii) REGISTER is the term commonly used for language variation of a non-dialectal type; e.g.
differences between polite and familiar language; spoken and written language; scientific, religious,
legal language, etc.
(iv) On the classification of verbs in terms of their relation to other ele-ments in the clause, see
Greenbaum and Quirk (1990), sections 10.1– 10.18 and 16.11–16.37. This aspect of lexical choice is
closely bound up with semantic relations between noun phrases in the clause: these have been
investigated by Fillmore3 under the heading of ‘case’, and by Halliday4 under the heading of
‘transitivity’. Their role in style is discussed in section 6.1. Factive verbs presuppose the truth of what is
being asserted (e.g. ‘Mary liked the show’). Counterfactives presup-pose the negation of what is
asserted (e.g. ‘Mary pretended to like the show’) and nonfactives leave the question of truth open (e.g.
‘I believe that Mary liked the show’).
(v) The traditional classification of adverbs and adverbials into adverbs of time, place, manner,
frequency, etc. is serviceable enough; a more thorough and systematic classification of adverbs is given
in Greenbaum and Quirk (1990) Chapters 7 and 8, where a major distinction is made between adjuncts,
disjuncts and conjuncts.
(vi) The delaying of the main ‘information point’ of a sentence by anti-cipatory and parenthetic
structure is discussed further in section 7.5. This is the defining feature of the traditional rhetorical
category of ‘periodic’ sentence, often contrasted with the ‘loose’ sentence.
(ix) Of course, the same word form may occur in more than one word class. For example, that is a
determiner (specifically, a demonstrative determiner) in ‘That day nothing happened’, a pronoun in ‘I
know that’, and a conjunction in ‘I know that he’s wrong’. In English, the overlap between the pronoun
and determiner classes, for instance, is very striking
3.3 Joseph Conrad: example 1
In the remainder of this chapter we apply the categories in section 3.1 selectively to three texts which
are comparable both in length and in that each is the opening passage of a short story. Our three
authors are Conrad, Lawrence and James. The procedure in each case will be to begin with some general
first impression of the passage, and then to make selective use of the checklist in order to bring to
readers’ attention what appear to be the most significant style markers of each. These style markers, in
turn, will be related to other style markers within the context of the passage’s literary function. In
section 3.7, we give a table of quantitative data (Table 3.1) from these three passages, so that our
analysis can be seen to be based on ‘hard evidence’. Cross-references to this appendix are given by italic
numbers (e.g. 16). From Joseph Conrad, The Secret Sharer:
On my right hand there were lines of fishing-stakes resembling a mys-terious system of half-submerged
bamboo fences, incomprehensible in its division of the domain of tropical fishes, and crazy of aspect as if
abandoned forever by some nomad tribe of fishermen now gone to the other end of the ocean; for
there was no sign of human habitation as far as the eye could reach
Our first impression of this passage is of a meticulously detailed setting of the scene for the story. The
description is clearly etched, so that we can reconstruct, in our mind’s eye, the whole topography. But
more than this, we have a vivid sense of the loneliness of the human observer, set apart from his
surroundings, and of ‘a mind energetically stretching to subdue a dazzling experience outside the self, in
a way that has innumerable counterparts elsewhere in Conrad’.
