In this paper I would like to present some observations on the intonation of radio newsreaders.
The observations are based on an auditory and instrumental study of 27 minutes of newsreading speech by 11 speakers, from the ABC and from the commercial stations 2GB and 2CH in Sydney. I also studied 12 minutes of non announcing speech from the same speakers, as well as 68 minutes of 'live' commercials, disc jockey chat, 'fine music' announcements and public information announcements from these and other speakers, so that, from time to time, I will be able to invoke a comparison with other types of announcing speech, or with non-announcing speech1.
Accent and the impartiality of the news
The first of my observations concerns accent — the term I use for what, elsewhere, is often called Sentence stress': the perceivable prominence of some syllables over other, adjacent syllables in actual connected speech.
Insofar as the literature on this subject sets out to provide rules to predict the placement of accents, one rule, the rule that 'content words' will receive an accent, is invariably present, although exceptions — ail too easily observed — are usually acknowledged. Thus content words may not be accented as a result of rhythmic adjustment, or because they embody information which is 'given' in the context, or because they are part of certain specific syntactic structures or of a 'contrastive intonation', to list only the most commonly mentioned exceptions2.
Given this basic accent rule, we can ask three questions about the way newsreaders accent their speech. In the first place, how often is the rule violated, or, how great is the number of exceptions? In the second place, how is the rule violated, or, what kind of exceptions are the most common? And, finally, how can we interpret the results or why do newsreaders accent the way they do?
To begin with the first question, the quantitative question, analysis shows that the non announcing speech of radio announcers contains
84 Aust. J. Cultural Studies, 2:1 (1984)
significantly more exceptions to the 'content word' accent rule than their announcing speech, but that there is, between the various genres of announcing speech, relatively little variation3:
Type of announcement
Percentage of accented single-stress content words
Newsreading Information Fine music announcing Popular Music announcing Commercials
Non announcing speech
94.9 94.3 93.75 93.36 92.39
85.81
table 1: The percentage of single stress content words which receives an accent in different types of announcement.
If what many writers on intonation have argued is true, that speakers accent those words which they feel to be important4; and if we accept that not every word issuing from the mouth of a radio announcer is worth its weight in gold; if we realize, also, that there are, between these different types of announcement, considerable differences in informational density (compare, for example, newsbulletins with the disc-jockey's recycling of the same old top 40 songs); then we can also accept, I think, that announcers, in general, are in the habit of making everything they say sound important, regardless of whether it is in any real sense, or for any other reason than that it is heard by a very large number of people.
Turning now to the second question, we find that two phenomena, two kinds of exception, account for almost all the instances in which single stress content words fail to receive an accent in newsreading — indeed, they account also for exceptions to other accent rules, for the placement of accents on prepositions, articles and auxiliaries, for instance, and even for the placement of accents on lexically unstressed syllables5.
(a) Rhythmic regularization
The first of these two phenomena is a tendency towards rhythmic regularity, that is, a preference for rhythmic feet of three syllables, and an attempt to avoid variation in the number of syllables per rhythmic foot, an attempt, in other words, to continue for as long as possible a regular disyllabic or, preferably, trisyllabic footstructure6.
85 Aust. J. Cultural Studies, 2:1 (1984)
Frequently newsreaders sacrifice information value to this kind of rhythmic regularity. All five ABC newsreaders I recorded, for example, accented the first sentence of this news item as follows (the accented syllables are underlined, and the rhythmic feet separated by single oblique strokes):
[1 [ 2 [ the con/1tainnership//] 1[2Asian Re/1nown//] [is/ 2due to leave/2Brisbane//] [ 2early this/1 afternoon//] [with a con/ signment of u/2ranium/1yellowcake// ]
In other words, although the verb 'leave', by any reasonable assessment, should be accorded some importance, some information value, it remains unaccented for the sake of rhythmic regularity.
Similarly, 3 out of the 5 ABC announcers failed to place an accent on 'forty' in the number name sequence 'forty one thousand eight hundred dollars' — again in order to continue, for as long as possible, a series of trisyllabic rhythmic feet:
[ forty/ 1one thousand//] [ 1eight hundred/2dollars// ]
In other types of announcing speech this kind of rhythmic regulari-zation also occurs, but there it remains localized, and tied to the content of the stretch of speech on which it is superimposed. Regularized monosyllabic rhythm is used, for example, to intonationally underline the importance of the title of a song in a disc-jockey announcement, or the address of an advertiser in a commercial. Or a more or less regularized series of rhythmic feet with three or four syllables per foot may be used to skip lightly over a less important, more or less parenthetic part of an utterance, for example in the speech of 2SM announcers, and in non announcing speech. But it is only in newsreading that rhythmic regularization becomes an overall characteristic of the genre, and bears little or no relation to the importance or lack of importance of the stretch of speech on which it is executed.
(b) the punchline phenomenon
The second phenomenon which causes newsreaders to violate the accent rules of English is an increase in the rate of accentuation towards the end of each news item — regardless of whether the words on which this increase is executed contain important information: the final intonation group in the item gets the punches, whether or not it is in fact a punchline. And this can even cause lexically unstressed syllables to receive an accent (there were 20 instances of this in my corpus of newsreading):
[Aus/tralia's/1highest/2score//] [in the/2three/ world/
1cups//]
[ in/2which they/ have par/1tid/ pated//] ] ]
86 Aust. J. Cultural Studies, 2:1 (1984)
These two habits, rhythmic regularization and the 'punchline' habit, can be observed in the speech of all newsreaders. And yet, members of the announcing profession have, at times, commented on them adversely. Thus Elwyn Evans (1977: 50), a former head of radio training at the BBC, has written this about rhythmic regularization:
. . . newscasters sound monotonous simply because their speech rhythms are too regular...
And for 'punchline' accentuation he blamed the individual newsreaders:
. . . reading out the phrases rather than concentrating on the sense is (...) the surest way to develop irritating vocal mannerisms, such as the piledriving thump which some newscasters bestow on the last words of every sentence, whether important or unimportant. For instance 'The pound had a better day, and share prices went up in the city'. Where else?. ..(49)
And this leads us to our third question, the question of interpretation. Is Evans' explanation satisfactory? Is it correct to ascribe to the lapses of individual announcers what is so evidently part of the practice of all newsreaders, not only in Australia, but also in other English speaking countries?
Let us review what we have found so far. We have established that accentuation, though largely governed by the accent rules of the English language, nevertheless leaves the speaker a margin of intona-tional freedom, and we have, for one particular accent rule, quantified this margin. We have established also that newreaders allow this margin of accentual freedom to be governed by further rules, the accent rules of professional newsreading, and that these rules are very formal, almost mechanical rules, quite unconnected to the importance of what is being said for the message as a whole.
Why should this be so? Why should newsreaders fail to use accentuation the way it was intended to be used — for emphasis, for creating contrasts, for discriminating between what is important and what is not? If one asks the newsreaders themselves, the answer is always the same:
. . . News demands first that there is impartiality. Therefore there is a kind of formality about the way you read it. You can't be moved by what you read. If you did that (... ) you'd be a hopeless newsreader, because you would be exciting prejudice (... ) or show your bias...
(ABC newsreader)
The accentual habits of newsreaders, it seems, serve to signify the impartiality of the news, to demonstrate, even if it has to be at the cost of intelligibility, that newsreaders are not themselves involved in the message they transmit, do not themselves discriminate between what is important and what is not. The newsreader's speech, in order to be impersonal and impartial, must become as devoid of inflection
87 Aust. J. Cultural Studies, 2:1 (1984)
as is official writing, negating the presence of the speaker by allowing itself to be subjected to formal, mechanical rules. 'Not concentrating on the sense' is not a lapse of individual newsreaders, but, on the contrary, an essential feature of newsreading speech, one of those features through which the mass medium, the institution which administers the news for us, symbolically expresses the values on which its credibility is founded.
Other aspects of newsreading accentuation point in the same direction. The distinction between 'given' and 'new', for example, is systematically ignored in newsreading — by the same announcers who do not ignore it in, for example, their fine music announcements or information announcements, or in their non-announcing speech. In this section from an ABC news item, the word 'Israeli' is accented, every single time it occurs, as a new and important item of information:
1 [ 1 [ 1 [ re/ports//] 2[are con/1tinuing to/2come in//] [of a/2big Is/lraeli/ push//] [over the/1border//] [into/2Southern/1 Lebanon//]] 1 [ 2 [ an Is/2raeli/Army com/ muinique//] [ con/firmed// ] 1 [ that Is/2raeli/1troops//11 [ had/2crossed the/1border// ] ]
In non-announcing speech, on the other hand, words are usually left unaccented after they have occurred once in the text, as is the case with the word 'news' in the following example — and note also how much greater the variation in number of syllables per rhythmic foot is here, and how this variation corresponds closely to the informational importance of the intonation groups for the message as a whole:
[ I've read/2news at all/1 sorts of/ places and the the// ]
[2blokes that/ write the news/1here//] [ 1em// ] [Certainly// ]
[ far/1 easier to/2read//] [and/1 flows/2better// ] [and that's/
not just because I'm/2working here/1 now// ] [ than/1any
news I've/ read anywhere/ else// ]
The case of attributions might also be mentioned. These occur a great deal in news, and one could argue that, for this reason, the verbs of attribution ('say', 'confirm', 'report',) might, in the context of news broadcasts, be taken for granted. Yet, in phrases like 'An ANL spokesman says', or 'a Palestinian spokesman in Beirut says', the verb 'say' is almost always accented. In this case the importance of the word does in fact tend to override the importance of rhythmic regularity, which betrays, I feel, something of the priorities of the newsreader: predictable as they are, attributions must be foreground-
88 Aust. J. Cultural Studies, 2:1 (1984)
ed because the newsreader's impartiality, the fact that the newsreaders themselves are not implicated in the message, must be foregrounded, continually and emphatically.
Juncture and the factuality of the news
A second set of observations I would like to make relates to juncture, that is, to the boundaries by means of which intonation divides the stream of speech into separate phrases.
In this case, too, I want to depart from the linguistic norm, from 'unmarked' intonation, to then investigate how often, and in which ways the linguistic norm is transgressed, and what we can learn about newsreading speech from a close study of these transgressions.
But what exactly is the linguistic norm? To which degree can the placement of intonation boundaries be predicted by syntax-based rules? What form should these rules take? On questions like these there is much divergence of opinion7.
In the present paper I will depart from the 'normative framework' of syntax-based boundary placement rules outlined by Crystal (1975), in my view the most satisfying and delicate treatment of the
problem so far. According to Crystal, every 'simple clause' is given 'international identity': it is preceded and followed by an intonation boundary, for example a pause, or pitchturn, or a lengthening of the
final syllable or syllables. The 'simple clause' he defines (1975: 16) as consisting
. . . maximally, of the elements Subject + Verb + Complement and/or Object with one optional Adverb, in this order, if each of the elements S, 0, or A is expounded by a simple nominal group...
In defining the 'simple nominal group', he deviates somewhat from Quirk et al. (1972), for whom the simple nominal group either consists of a noun modified only by a determiner, or is realized solely
by a pronoun or proper name, and he includes also what Quirk et al call the 'type i complex group', a group with
. . . modification consisting of one adjectival premodifier and/or one prepositional phrase postmodifier... (16)
These groups, says Crystal, behave intonationally in the same way as 'simple groups'.
Crystal gives, of course, many further refinements, to cope with ellipsis in coordinate clauses, for example, and with such special cases as reported speech, 'comment clauses', tag utterances, but for reasons of space I will not discuss these here.
Crystal's rules permit the identification of different modes of marked juncture, two of which I will discuss here: the disjoining of elements of the simple clause, and the disjoining of elements of the simple nominal group. In news, as can be seen from the following
89 Aust. J. Cultural Studies, 2:1 (1984)
tables, both simple clauses and simple nominal groups are frequently further broken up into intonation groups8.
Type of announcement
Newsreading
Fine music announcing
Information
Commercials
Popular music announcing
Mean percentage of junctures intonationally
breaking up simple clauses
38.8
37.2
32.4
27.4
25.4
Non announcing speech
22.5
table 2: Mean percentage of junctures intonationally breaking up simple clauses in ways incompatible with Crystal's normative framework. Comparison between different types of announcement.
Type of announcement
Information
Newsreading
Fine music announcing
Commercials
Popular music announcing
Mean percentage of
junctures intonationally
breaking up simple nominal groups
12.5
14.4
14.4
14.6
20.0
Non announcing speech
17.2
table 3: Mean percentage of junctures intonationally breaking up simple nominal groups in ways incompatible with Crystal's normative framework. Comparison between different types of announcement.
In which ways do newreaders deviate from the norms? Looking first at the disjoining of elements of the simple clause, we find that the grammatical subject is the most frequently disjoined element (55.7 °o of cases). An example, from a 2CH bulletin (the junctures are indicated by a double oblique stroke (//) and the intonation groups are enclosed in square brackets):
[[ Debra//] [ was/ wearing a/ blue/1poloneck/2jumper// ] 2 [ a/ white/2cheese/ cloth/1 skirt// ] 1 [ and/2clog/ type/ 1shoes//] [with/2red/straps//]]
90 Aust. J. Cultural Studies, 2:1 (1984)
But other components of the simple clause, too, are made into separate groups, even if this leads, at times, to very short intonation groups. From a 2GB bulletin:
[l[l Lifesavers//] [ 1used//] [2special tech/1niques//] 2 [to/ keep/ Larkin a1/float in the/2water//] [ un/ till_ / medical/1help ar/2rived// ] ]
No surprise, perhaps, that one announcer, when discussing news-reading with me, illustrated his words with what is, in the context of non-announcing speech, a good example of expressive intonation. "In news", he said,
[2vou are/1simply//] [re/1lating//] [eh/1facts//] 1[1to people//] ]
To explain why these marked forms of juncture are so common in newsreading, it is useful to look at some of the assumptions that underlie the writing of news. News, according to the textbooks (Mencher, 1977: 112), is consciously written to a schema in which the reporter must, before starting to write, ask himself two questions:
... 1.. What was the most important thing that happened? 2. Who was involved —who did it or who said it? ...
His sentences, in order to incorporate the answers to these questions, should:
... begin with the subject, should be closely followed by an active verb, and should conclude with the object of the verb (... ) The S- V-0 construction is consistent with the thinking pattern of the reporter as he or she structures the lead (and) is the most direct way of answering the first two questions... (115)
And the same principle, according to a textbook on radio journalism by an Australian author, must guide the way newsreaders 'group their words' (cf. Herbert, 1976, p.94).
The 'who' and the 'what', are presented as separate, 'hard' facts, disconnected, rather than being brought in relation to each other. Again the speaker must follow a formal rule, rather than group together what he feels belongs together.
Turning now to the disjoining of elements of the simple nominal group, we find that this mode of disjoining occurs, invariably, in order to treat pre- or postmodifying elements as separate information units, especially when these modifiers contain the type of information which serves, more than anything else, to highlight the factuality of the news: numbers, times, locations, the age and place of domicile of people in the news, and so on. An example, from a 2CH bulletin:
91 Aust. J. Cultural Studies, 2:1 (1984)
[ [2One of the /1fishermen//] [2forty/1fiveyear/old// ] [2James/1Browning// ] [of/2Brunswick/1 Heads// ]
As a result of this habit, elements of the nominal group are intona-tionally treated as though they were elements of the clause, postmodifiers intonationally disconnected from the nouns they modify, for example, and pronounced as separate, rather than modifying items of information, as is 'over the border' in this example, quoted earlier:
1 [1[re/1ports//] [are/con/1tinuing to/2come in//] [of a/ 2big Is/2raeli/ push// ] [over the /1border//] [into/2Southern/ lLebanon//] ]
In this way phonological disconnection serves to signify a lack of connection between events, makes it possible for the news to circumvent the necessarily political nature of relating events to each other, to present facts as isolated, things-in-themselves, and to deliberately leave unused the potential of intonational grouping to relate the part of speech together on the basis of the speakers' judgments of what they feel belong together. Thus the news signifies its 'independent event epistemology', an epistemology in which:
... society is merely an aggregate of individuals ( ... ) permanence is lacking ( ... ) structures tend to decay ( ... ) each question has its answer, unrelated to others...
(Maruyama, 1980: 34-35)
Phonological continuity and the authority of the news
A third set of observations relates to what I have called 'continuity': an intonation group is either 'open' (displays continuity) — and this is often expressed by a rising or level terminal pitch, and usually indicates that the utterance is incomplete — or it is 'closed' (displays finality) — and this is normally signalled by a falling terminal pitch, and, as a rule, indicates that the utterance is complete9.
The linguistic norm, in this case, is that intonation groups are 'open' when they are sequence-initial or —medial, and 'closed' when they are sequence-final. Intonation sequences are, so to speak, the intonational equivalence of sentences, series of intonation groups (enclosed, in the transcriptions in double square brackets) separated from each other by an intonation boundary which is perceptually more conspicuous than the boundaries between the intonation groups within the sequence10.
But this norm, too, is, at times, transgressed by newsreaders, as the following table shows11.
92 Aust. J. Cultural Studies, 2:1 (1984)
Type of announcement
Popular music announcing Commercials
Information
Newsreading
Fine music announcing
Mean percentage of sequence-initial and -medial closed boundaries
12.3
7.4
6.9
3.2
2.8
Non announcing speech
6.1
table 4: Mean percentage of sequence-initial and -medial closed groups in different types of announcement.
The occurrence of closed sequence-initial and sequence-medial groups can be accounted for by three distinct intonational habits:
(a) the headline effect
A title or name, incorporated in the grammatical subject of the opening sentence of a news item, is frequently given a closed group ending, so that a kind of headline effect is created (an acute accent on the final syllable of an intonation group indicates that the group is open, a grave accent that it is closed):
[ [l [ the/ value of the Aus/2tralian/1dollar//] [ 2has/ gone/ 1down//] [to its/2lowest/1level//]
(b) the utterance ending
Towards the end of a bulletin, or, occasionally, a non-final item within a bulletin, not only the final intonation group, but also two or three of the preceding groups may be closed, to announce in advance, as it were, that the end of the utterance is near. This effect is usually accompanied by discrete step-downs in pitch and loudness over the final closed groups:
[ 2 [ Sydney/ trainer/2Theo/1Green//] [1says//] 1[1he's/not
sur/2prised//] [that/ 1stable/2lads are//] 2taking/1drugs//]]]
(c) the authoritative intonation
But newsreaders also all create further sequence-medial closed groups, at apparently random places in the script, as if fulfilling a requirement that a certain number of additional closed groups must be created, at points to be decided by the individual announcers:
93 Aust. J. Cultural Studies, 2:1 (1984)
[l [ 2[a/1twenty one year old/2man// ] [ is in in/2tensive/1 care//] [ in/ Sydney's/2Mona Vale/1 Hospital// ] 1[after/2breaking his/ 1back//] [in an/2accident at a/ beachside/1 swimming pool//]
It is a habit which not only further enhances the lack of connection between individual information units, the self-sufficient definiteness of each of the facts reported, but also adds a sense of authority to newsreading. The falling pitch is assertive, the tone of definitive statements and commands. But if the 'completeness' and 'definiteness' of the closed intonation is not accompanied by some form of semantic completeness, some definiteness in the verbal content, it becomes a random definiteness, assertiveness for its own sake, a deliberate foregrounding of the fact that the speaker, in radio, always has the final word.
The same use of the closed group can be observed also in other genres of radio announcing, especially in commercials. But here the distribution is not random. Here the marked group endings occur at keypoints in the utterance, at privileged moments in the rhetorical structure of the commercial — to render separate the items from the range of products offered, for example, an important aspect of the persuasive strategy of the commercial, as it shows that 'something is provided for all, so that none may escape* (Adorno and Horkheimer, 1972,p.l23):
[Try their/ Peking/stvle/2fillet/1 steak/ ] l[ 2gr/1 delicate/ chicken and/ ham/1 rolls// ]
Or it adds further assertiveness to the actual injunction to buy the product:
lSo//] 2[2Nippon/lout/] 1[to/1Peter/2Williamson//] eh/2Liver/pool//]
It is only in newsreading that the distribution of authoritative intonations bears no relation to the meaning of the message, only in news-reading that it becomes an overall-genre characteristic, displayed for its own sake.
Rate of utterance and the immediacy of the news
Finally, news, in comparison with other types of announcement, is characterized by an extremely high rate of utterance12:
94 Aust. J. Cultural Studies, 2:1 (1984)
Type of announcement
Rate of utterance (in
syllables per second)
Newsreading
Information
Popular music announcing Commercials
Fine music announcing
4.69 4.59 4.48 4.1 3.93
Non announcing speech
3.89
table 5: Mean rate of utterance (in syallables per second) in different types of announcement.
This great phonological hurry of the news serves, I believe, to signify its immediacy and recency, and illustrates what Raymond Williams has said in his study of the flow of television programmes (1974:116):
... over much of the actual newsreporting there is a sense of hurried blur. The pace and the style of the newscast take precedence over the items in it. This sense of hurried transmission from all points is then in sharp contrast with the cool deliberation of the commercials. The flow of hurried items establishes a sense of the world: of surprising and miscellaneous events coming, tumbling over each other, from all sides...
Conclusion
I have attempted to sketch how what I would c^all professional codes of intonation settle themselves in the margin of intonational freedom which the English language provides — professional codes which serve to express the values that underlie the practice of news-gathering and reporting, to signify — continually and overtly — the impartiality and the formality of the news, to represent facts as self-sufficient, separate, hard, to stress its own authority, to express its own immediacy and recency.
These professional codes are not arbitrary codes, but expressive, iconic codes: intonation is, or rather, resembles what it signifies. Lack of phonological contrast (contrast in phonological importance) serves to signify the absence of informational contrast. Phonological disconnection serves to signify the disconnection of facts and events. The intonation of the assertion, when used arbitrarily and randomly, serves to endow newsreading with an overall assertiveness and sense of authority. Phonological hurry signifies the fever with which news is gathered and transmitted from all over the world.
95 Aust. J. Cultural Studies, 2:1 (1984)
Perhaps we have hit here upon a crucial difference between ritualized speech such as radio announcing, and individual expressiveness, the true exercise of one's intonational freedom: in ritualized speech the iconic signifiers become overall stylistic characteristics, expressing what is permanent in some form of socio-culturally determined interaction. In individual expressiveness they are related to the sense of the message: add, for example, phonological hurry to a part of a text which, one way or another, deals with hurry or with something that happens in a hurry; add phonological lack of contrast, or disconnection, or authority to parts of texts which are about disconnection, or lack of contrast, or authority .
This is not to say that individual expressiveness is altogether absent from radio announcing. There are, for example, information announcements in which the information pointing capacity of intonation is used discerningly, to separate the important from the* unimportant, the given from the new, for example ABC fine music and information announcements. But in newreading it is absent. Here, to continue the same example, information pointing is turned into an overall stylistic characteristic rather than relating to the information conveyed. In newsreading it expresses only the abundance of new and important information which, so the medium would like us to believe, characterizes the news with which we are provided. And these overall stylistic characteristics realize the professional codes of intonation I have attempted to describe in this paper.
Theo van Leeuwen teaches at Macquarie University.
96 Aust. J. Cultural Studies, 2:1 (1984)
Notes
(1)A detailed description of the corpus and the methods of analysis used
can be found in Van Leeuwen (1982), ch. 1 and 2.
(2) Cf. Van Leeuwen (1982), pp.34-48 for a full discussion of the problem of
accent rules.
(3) As content words are regarded: nouns (including proper names), verbs
(except auxiliaries and the verb 'be' used as a mere link), adjectives and
adverbs (with the exception of conjunctions).
Only single stress content words are included in the table: if these do
not receive an accent, they can reasonably be said to have been treated
as fully unimportant, whereas words with two or more lexical stresses
which receive only one accent retain at least some of the importance
which theoretically (grammatically) accompanies their status as a content
word. Van Leeuwen (1982) discusses also the destressing of words with two
or more lexical stresses, cf. pp.147-153.
The variation between different announcing genres in table 1 is not statistically
significant, but variance between announcing speech and non announcing speech
is: t = 3.58; p 0.05; d.f. between 5; d.f. within 56). Correction for unequal sample
sizes was made according to Hays (1963),p.322.
(4) Cf., e.g., Bolinger (1961; 1964; 1972); Trim (1964); Cruttenden (1970);
Ladefoged (1972); Crystal (1975).
(5) Van Leeuwen (1982) deals quantitatively also with the degree to which
accent rules other than the 'content word' rule are violated in newsread
ing and other genres of announcing speech, cf. pp. 153-174.
(6) Cf. Van Leeuwen (1982), pp.124-133, for fuller discussion, and statistics
on the incidence of disyllabic and trisyllabic rhythmic feet in various
genres of announcing speech.
(7) Cf. discussion in Van Leeuwen (1982), pp.49-52.
(8) The figures in tables 2 and 3 are percentages of the total number of
'juncture points', that is, of the sum of the junctures predicted by
Crystal, and the number of actually realized junctures not predicted by
Crystal. Variance is significant in both cases: for table 2 F = 6.67; p
0.05; for table 3 F = 2.79; p 0.05; d.f. between 5; d.f. within 56.
(9) I prefer not to tie the notion of continuity too closely to pitch, as I have
observed other phonetic cues of continuity, and as experiments showed
pitch movement to be a poor criterion for perceiving whether a group is
open or closed (cf. Van Leeuwen (1982), pp.67-75).
(10) Cf. Van Leeuwen (1982), pp.75-84, for more discussion of higher order
intonational junctures.
(11) Variance significant: F = 3.27; p 0.05; d.f. between 5; d.f. within 56.
(12) Variance significant: F = 5.13; p 0.05; d.f. between 5; d.f. within 56.
(13) This kind of expressiveness is well illustrated in Fonagy's analysis of
poetry readings (1976).
97 Aust. J. Cultural Studies, 2:1 (1984)
References
Adorno, T.W. and M. Horkheimer, (1972). Enlightenment as Mass Deception, (New York: Herder and Herder).
Bolinger, D.L., (1961). 'Contrastive accent and contrastive stress' Language 37:83-96.
(1964). 'Around the Edge of Language: Intonation', in D.L. Bolinger, ed. Intonation, (Harmondsworth: Penguin, 1972).
(1972). 'Accent is predictable (if you're a mindreader)' Language 48: 633-644.
Cruttenden, A., (1970). 'On the so-called grammatical function of intonation', Phonetica2\\ 129-137.
Crystal, D., (1975). The English Tone of Voice, (London: Edward Arnold).
Evans, E., (1977). Radio — a Guide to Broadcasting Techniques, (London: Barrie and Jenkins).
Fonagy, I., 11976). The Voice of the Poet', in A. Makkai, ed. Toward a
Theory of Context in Linguistics and Literature, (The Hague: Mouton,
1976). Hays, W.L., (1963). Statistics for Psychologists, (New York: Holt, Rinehart and
Winston). Herbert, J., (1976). The Technique of Radio Journalism, (Melbourne: Edward
Arnold, Australia). Ladefoged, P., (1972). A Course in Phonetics, (New York: Harcourt, Brace,
Jovanovich). Maruyama, M., (1980). 'Information and Communication in Poly-
Epistemological Systems', in K. Woodward, ed. The Myths of
Information, (London: Routledge and Kegan Paul).
Mencher, M., (1977). News Reporting and Writing, (Dubuque: Wm. C. Brown).
Quirk, R., S. Greenbaum, G.N. Leech and J. Svartvik, (1972). A Grammar of Contemporary English, (London: Longman).
Trim, J.L.M., (1964). Tenetic Stressmarks in German', in D. Abercrombie,
ed. In Honour of Daniel Jones, (London: Longman). Van Leeuwen, Th. J., (1982). Professional Speech: Accentual and Junctural Style
in Radio Announcing, unpubl. MA thesis, Macquarie University, Sydney. Williams, R., (1974). Television, Technology and Cultural Form, (London:
Fontana).
New: 21 October, 2019 | Now: 13 November, 2019