Word Stress and Sentence Accent PDF

# Word Stress and Sentence Accent ## Héctor Ortiz-Lira, PhD * Lecturer in English and Spanish Phonetics, Universidad Metropolitana de Ciencias de la Educación, Universidad de Santiago de Chile ## Presentation The materials contained in this booklet have been designed to meet the needs of the stu...

# Word Stress and Sentence Accent ## Héctor Ortiz-Lira, PhD * Lecturer in English and Spanish Phonetics, Universidad Metropolitana de Ciencias de la Educación, Universidad de Santiago de Chile ## Presentation The materials contained in this booklet have been designed to meet the needs of the students who are taking the last courses in English phonetics at teacher-training college level. We intend to offer the latest theory available in a way which is accessible to students, together with plenty of practice material for marking and reading. This booklet is divided into four main sections: 1. An analysis is carried out of the basic notions related to prominence. 2. The problem of stress in simple words is revisited. 3. The topic of stress in compounds is addressed. 4. Postlexical (or sentence) accentuation is examined in detail. We are aware that the students' success in matters of prosody is particularly influenced by their mother tongue, and that is why constant comparison with Spanish is made, in an effort to point out differences and similarities. A final appendix contains practice material for users to analyse and mark. This, together with the examples in the chapters, represents the largest collection of data on the subject ever published for EFL purposes. A key to the exercises together with a cassette containing the majority of the examples in the text are also available. ## 1. Fundamental Concepts of the Theory of Prominence ### 1.1. Introduction In this opening chapter, we shall look at various concepts which are central to a theory of prominence. This includes an analysis of notions such as stress, accent and prominence, a discussion of the thorny issue of degrees of stress, notation systems for stress and accent, and finally a re-examination of stress shift. Since we are catering for learners whose mother tongue is Spanish, constant reference to similar problems in that language is included in our discussion. ### 1.2. Prominence The first task we must embark upon is to elucidate the distinction between the notions responsible for highlighting portions of utterance. We shall begin by analysing the most basic of the concepts, that of prominence. When we speak, we give more emphasis to some parts of an utterance than to others. We can make a syllable stand out with respect to its neighbouring syllables in a word, and in doing so, the word containing that particular syllable will stand out with respect to the rest of the words in the utterance. The elements which produce prominence at syllable level are: 1. Pitch contrast 2. Loudness 3. Length 4. Quality As with all aspects of speech, these correlates can be described and defined in articulatory, acoustic, and auditory terms, i.e. from the point of view of the person (speaker) who produces prominence, from the physical viewpoint (that is, sound wave behaviour), and from the listener's angle (i.e. the way he perceives prominence). In the EFL situation, acoustic criteria seem to be of limited application and, consequently, we shall concentrate on the production and, more particularly, on the reception stages, rather than on the transmission phase. ### 1.2.1. Pitch Differences in pitch are produced at the vocal folds: the tenser the vocal folds, the faster they vibrate, and the higher the note that is produced-un articulatory characterisation. But, we cannot actually feel our vocal folds vibrating faster or slower. What we hear is a higher or lower note. So articulatorily pitch depends mainly on the tension and consequent rate of vibration of the vocal folds. Auditorily, it is that property of a sound, in terms of which it can be placed on a scale running from high to low. ### 1.2.2. Loudness From an articulatory point of view, loudness is caused by greater muscular energy and breath force. Auditorily, it is defined as that property of a sound which enables us, using only our ears, to place it on a scale going from loud to soft. So the speaker feels this feature in one way — extra energy— and the listener hears it in another way — extra loudness. ### 1.2.3. Length It is not necessary for our purposes to define length from the point of view of the person who produces it. Auditorily, it is that property of a sound which enables us, using only our ears, to place it on a scale that goes from long to short. ### 1.2.4. Quality Articulatorily, quality depends on the shape of the cavities or resonators (mainly the mouth), whose function is to modify— sometimes amplify, sometimes suppress— the almost inaudible note produced at the vocal folds (commonly known as ‘voice’ but more appropriately, vocal fold vibration). Auditorily, quality is defined as that feature in terms of which two sounds, similarly presented and having the same pitch, loudness and length, are still perceived as different; we say that [æ] and [u:], for instance, differ in terms of quality. In English, syllables containing [ǝ], [i], [u] and syllabic consonants are normally perceived as weak; two vowels, [1] and [u], play a duel role, and the rest are strong. Another way of referring to these two types of vowels is full and reduced. The following table summarizes the auditory correlates of prominence, c.g. | Correlates | Syllables | |-----------|------------| | Pitch | high ++ low | | Loudness | loud ++ soft | | Length | long ++ short | | Quality | strong + weak | If we analyse the title of this chapter from the point of view of prominence, we will most probably conclude that there are four syllables which stand out from the rest. Here, we show them in small capitals, c.g. FUNdamental CONcepts of the THEOry of PROMinence. These four syllables are made prominent mainly because they contain strong vowel sounds produced with relatively greater loudness and, more important, with pitch movement or contrast. In sum, all four elements can play a part in making a syllable stand out over the rest. They do not all play an equally important part, though nor are all four always present together. In the example above there are also eight non-prominent syllables. These are, from left to right, -da-, -tal, of, the, -ry, of, -in-, -ence, all of which consist of weak vowels or, in the case of -tal, absence of vowel. There are two syllables which are more prominent than the eight weak ones, but less prominent than the first four we analysed, c.g. -men- and -cepts. These two contain the strong vowel [e], and it is fairly possible that in a normal reading of the utterance, a native speaker would identify syllable-men- with some kind of a ‘rhythmic beat’, i.e. a combination of extra loudness and length. The really prominent syllables are so identified because they display the most decisive of the elements producing prominence — pitch contrast. There is a notation system called ‘interlinear tonetic’, which is a type of narrow transcription for intonation in which each syllable is represented by a dot; a large dot indicates a prominent syllable; a small dot, a non-prominent one. Since the upper line corresponds to the speaker’s top pitch range and the bottom line to his bottom pitch range, it follows that the pitch of each individual syllable, prominent and non-prominent, can be shown. In the following diagram, the size and hight of each syllable indicator represent only relative values, e.g. ``` (1.1) fandəmentl konsepts əv də Ərəri əv promınəns ``` ### 1.3. Stress and Accent The labels stress and accent have a long tradition in prosodic studies, but it is only since the beginning of the seventics that the majority of phoneticians have come to agree on the notions they refer to. Whereas prominence is exclusively a phonetic notion ‘used to refer to the general distinctiveness of an utterance of any length’ (Crystal, 1969: 120), stress and accent have phonological status. In this booklet, we are also using prominence as a blanket term which covers all phenomena related to the highlighting of syllables in words and of words in longer utterances; hence the title of the chapter. Important differences can be established between stress and accent. In the first place, stress is a feature of the word, a lexical abstraction or a decontextualised form which becomes concrete realization only if the word gets an accent in an utterance; hence, accent is a feature of the utterance, i.e. a concrete, contextualised category. This makes accent an observable phenomenon, as opposed to stress, which is merely analytical. To sum up, stress is a potential for accent and accent is an exponent of stress. Let us explain the above in practical terms. Dictionaries give every word of two or more syllables a stress mark; each stress pattern represents not only an entry in the dictionary, but also an entry in our mental lexicon, i.e. each word is stored in our mind in the form of a pattern of sounds (i.e. a phonemic pattern) and a pattern of prominences (i.e. a stress pattern); in more technical terms, stress is a phonological property of words. Stresses in words, however, cannot guarantee that the word will get an accent in a particular context, since this will depend on two main factors — the rhythmic structure of the utterance, as will be explained in §1.7— and pragmatic-discoursal principles, which regulate accentuation according to: 1. Word classes—content words are much more accentable than structural words. 2. The information structure of the utterance (roughly, words representing new information get an accent while those conveying given information tend not to). 3. Language specific rules which, in the case of English, assign more accents to nouns than to verbs in particular constructions. (See chapter 4.) Monosyllables are never shown bearing stresses in dictionaries. This does not mean that they do not have a stress, but simply that showing the stress on the only syllable available would be redundant. This explains why an unstressed dictionary form may very well need an accent in a given context, as is the case in STRESS and Accent have been NEAR SYNonyms for a LONG TIME, where a normal reading would assign accent to four monosyllables. The fact that words make use of pitch contrast in utterances in order to convey prominence — a feature they do not display at lexical, decontextualised level — has made various authors differentiate between stress and accent according to which of the correlates of prominence is the most dominant feature. Therefore, they define accent as a feature of prominence whose main perceptual component is pitch change; this has originated the term ‘pitch accent’, widely used in modern descriptions; by way of contrast, they define stress as a ſcature of prominence in which the dominant perceptual component is loudness. According to this view, in (1.1) the first three prominent syllables bear stress, and the last, accent. Followers of this view are A.C. Gimson (1962), Crystal (1969). Lehiste (1970), O’Connor & Arnold (1973), and Cruttenden (1997). ### 1.4. Accent and Prominence As was pointed out earlier, the most casily perceived type of prominence is pitch prominence signalled by changes in pitch; but since pitch contrast also happens to be the main indicator of accent, the two phenomena, accent and prominence, are often confused in certain contexts. In this section, we shall attempt to point out the danger implied and clear up this confusion. There are three main places in an intonation group where pitch prominence can be confused with pitch accent: 1. The first occurs just after the last (nuclear) accent, particularly if the prominent syllable is said on a relatively high pitch, e.g. some in example (1.2), or if it performs a rising pitch movement after an accented fall, e.g. does in (1.3). In interlinear notation, prominent syllables are marked as o, c.g. ``` (1.2) dju wont sam (1.3) bobi daz ``` Confusion between accent and prominence in this position is the most troublesome of all, since the major part of the meaning of the whole utterance is conveyed by the portion beginning at the last accent (technically known as ‘nuclear tone’). What is crucial in our theory is that neither of these prominent syllables is accented because they do not initiate pitch contrast, but simply complete it — some completes a rise and does, the rising end of a fall-rise. We shall say that a syllable is accented when it begins pitch contrast. As can be seen in the interlinear representation of examples (1.2) and (1.3) (contextualised below as (1.4) and (1.5), respectively), the last accents —WANT and BOB— are actually beginning pitch movement: the former, a movement ‘up-from’ and the latter, a movement ‘down-from’. Spanish learners often mistake prominent syllables for accented ones at the end of utterances, particularly when a rising tone is involved, which results in an unmistakable sign of foreign accent, cf. ``` (1.4) These grapes are dcLIcious. D'you WANT some? (*D'you want SOME?) (1.5) A: Nobody ever WRITES to her. B: BOBby does. (*Bobby DOES.) ``` Ambiguity between accent and prominence can also take place before the first accent in the intonation-group, in the stretch of utterance technically known as ‘pre-head’, and it involves words which are of the unaccentable type and usually take weak-form pronunciations, e.g. pronouns, auxiliary verbs, conjunctions, etc., as in ``` (1.6) it dzəst didnt əks: tə mi ``` Here, the syllables it just are prominent but unaccented. The first pitch accent is realized as a ‘step-down’ on did and the nuclear accent cur, as a movement ‘down-from’. In normal circumstances, the first two syllables would be very near the ‘baseline’, i.e. slightly above the bottom pitch. (This high, though unaccented, stretch of utterance conveys an emphatic effect). Examples such as There are NO possibilities! /ðərə nəu posəbılətiz/ and You'd've THOUGHT it was FEAsible! /jud av ost it wəz fi:zəbl/ can be said with a pitch pattern very similar to that in (1.6), particularly if a protesting overtone was to be attached; both can begin with two high (therefore prominent), unaccented syllables. Finally, ambiguity between prominence and accent can occur when the first pitch accent, which indicates the beginning of the ‘head’, and the last pitch accent, which is the nucleus, form a sequence of level tones, either high or low. The accentual status of any other prominent syllables occurring between these two accents will be obscured by the absence of pitch contrast, as is the case of the syllables ask and who’s in It’s NO good ASKing WHO’S to BLAME. Some authors refer to these syllables as ‘tertiary stresses’, since stresses do not involve pitch prominence. In the present analysis, we shall leave the question unsolved and refer to these syllables as either accented or prominent. Fortunately, in this position in the intonation-group, the neutralisation of the contrast does not have any practical effect, c.g. ``` (1.7) its nou gud aiskın huiz tə bleım ``` The pitch accents in a head such as that in example (1.1), sometimes referred to as ‘stepping head’, form a series of steps-down in which they do in fact initiate a movement and will be considered accents. In summary: 1. Not all prominent syllables are accented. 2. The majority of accented syllables are prominent, though — given the phonological status of accent it is perfectly possible to find accented syllables which are perceived as non-prominent and, in extreme cases, accented syllables which are hardly perceived at all, c.g. the gesture which realizes the first accented syllable in Jones’ much quoted utterance Thank you pronounced [kkju] (1956: 245). The formulaic overtone implied by syllable thank starting a rising movement from a rather low pitch makes the listener perceive only the second, prominent syllable, you. (Jones said there was a subjective stress on the first double consonant.) This solution poses another problem for the identification of accent, that of the ‘subjective effort’ hypothesis. The view is disputed by phoneticians who deny the possibility of considering syllables which are not perceived by the listener. We can think of similar patterns in Spanish, c.g. ¡Claro! as an answer implying the obvious, with a very low, voiceless first syllable. The interlincar notation of the English utterance shows a prominent syllable only, c.g. ``` (1.8) k kju ``` The native speaker and listener do not always depend on the same clues as the foreign student for the perception of accent. When two native speakers communicate, they are sharing the same linguistic code, and consequently will make judgements based not on what they actually hear, but on what they know they should hear. From this point of view, the native listener’s perception of accent is subjective as it is influenced by his previous knowledge of the language. He normally knows where accent is and will seldom confuse it with prominence. This type of confusion may lead to ambiguity. ### 1.5. Degrees of Word Stress The question of how many levels of stress can be distinguished in English words is not without problems. The same can be said of Spanish, for that matter. Spanish speakers are normally aware of the existence of one syllable in every word (of more than one syllable) which they identify as accented: this is the syllable carrying primary stress. The existence of other (secondary) stresses in relatively longer words is an aspect of Spanish phonology native speakers tend not to be aware of, and one on which linguists do not fully agree. The existence of at least two degrees of stress in certain English words of two or more syllables is an issue more firmly settled. The traditional American school of prosody accounts for four degrees — primary, secondary, tertiary and weak (e.g. Trager & Smith, 1951: 37). Chomsky & Halle (1968: 16) speak of primary, secondary, tertiary and quaternary or zero. In the British school, Kingdon (1958b) recognises the existence of three degrees — primary, secondary and weak or unstressed. The latest version (1989) of the IPA (International Phonetic Association) alphabet provides marks for two degrees only, primary and secondary. Of the two pronouncing dictionaries published in Britain in the 90’s, LPD (Wells, 1990: 683) recognizes three main degrees of stress (primary, secondary and tertiary) plus two further degrees, which are typical of syllables carrying no stress marks, but which have a strong vowel (fourth degree) and a weak vowel (fifth degree). Although EPD (Jones, Roach & Hartman, 1997: xii) recognizes three levels, it finally sticks to only two in order to avoid unnecessary complexity. Roach (1991) operates with three —primary, secondary and prominent syllables. Lastly, GPE (Gimson & Cruttenden, 1994) distinguishes four degrees —primary, secondary, full vowel and reduced vowel, but at citation level, however, it operates with only two marks, one for primary and one for secondary. The following table shows the various levels of word stress according to GPE (1994), LPD (1990) and the present author: | GPE | LPD | This work | |:------:|:------:|:----------:| | 1 | 1 | 1 | primary stress | | 2 | 2 | 2 | secondary stress | | 3 | 3 | 3 | prominent syllable | | 4 | 4 | 4 | non-prominent syllable | | | 5 | | | | | | | | Only two degrees of stress will be accounted for, and, consequently, marked in the present analysis. These two stresses have phonological status, i.e. their distribution in the word, together with the phonemic pattern - that is, each word taken as a sequence of phonemes and stresses - specify the phonological information with which the lexical item is recorded in the mental lexicon. The other two types of syllable, prominent and non-prominent, are mere phonetic categories and are left unmarked. A syllable is prominent when it contains a strong vowel or displays (but does not initiate) some kind of pitch movement or performs a rhythmic beat. If we want to establish a correlation between degrees of stress, as a purely lexical fcature, and accent, as a contextual feature, we can conclude that primary stresses are the most logical candidates for nucicar accents and that sccondary stresses can become nuclear accents only in utterances implying some kind of contrast. (Problems related to accentuation in contexts will be analysed in chapter 4.) This type of difference between primary and secondary stresses is, then, of a functional nature; another view would be to differentiate between them in strictly phonetic terms and say that one is stronger (or weaker) than the other. A characterisation in perceptual terms is, however, subjective and liable to lead us into difficulty. The following diagram shows the possible correlations, starting from the least marked (i.c. ‘normal’) contexts; means ‘can be realised as’, c.g. | Lexical level | Contextual level | |--------------|--------------------| | primary stress | nuclear accent, prenuclear accent, unaccented syllable in pre-head and tail; | | secondary stress | prenuclear accent, unaccented syllable in pre-head and tail, nuclear accent in contrastive contexts; | | prominent syllable | unaccented syllable in pre-head and tail, nuclear accent in very contrastive contexts; | | non-prominent syllable | unaccented syllable in pre-head and tail. | ### 1.6. Word stress notation Phoneticians have used different notation systems to represent word stress. Some of them show primary stresses only, while others show primary and secondary stresses; some notations are of a more phonological nature, whereas others, more phonetically biased, indicate the pitch movement which normally accompanies stressed syllables — usually a falling tone on the last, primary stress. The following foll are the most common types: 1. The syllable carrying primary stress is capitalised; the secondary stress is left unmarked, e.g. pronunciation. BAby sitter, plum PUDding 2. The stressed syllable is underlined; either the primary only or both, primary and secondary stresses, can be shown, e.g. pronunciation, pronunciation, baby sitter, plum pudding, plum pudding 3. The stressed syllables are assigned numbers, 1 for primary, 2 for secondary, etc.; other numbers can be used to indicate further degrees of stress, e.g. baby sitter, plum pudding, brand-new car, elevator operator 1 2 3 4 4. Stress marks are placed either before or on top of cach syllable bearing the stress; several sub-systems can be identified, among which the system that uses for primary and for secondary is the most widespread in EFL dictionaries and other reference books produced in the UK; these marks have been recommended by the IPA for over 70 years. The tonetic-stress marking system invented by Kingdon (1939) continues to be the most widely used in intonation courses produced in Britain, c.g. 1.6.4.1 (=primary,, =secondary,, =tertiary) decon, tami'nation, 'baby sitter, 'baby sitter, 'baby sitter 1.6.4.2 (=primary, '=secondary) 'decon'tamination, 'baby sitter, plum `pudding, 'plum `pudding 1.6.4.3 (=primary,' =secondary,, =tertiary) 'decontamination, `ele, vator 'ope,rator, 'black, board 'rubber 1.6.4.4 ('=primary, ^ = secondary, `= tertiary) dêcontàminátion, White House, white house (=not a brown one), élevator ôpcràtor 5. All syllables are accounted for in ‘interlincar notation’ — a sort of pictorial, narrow transcription indicating stresses/accents and pitches. The conventions are as follows: a large tailed dot = primary stress; a large filled dot = secondary stress; a large empty dot = prominent syllable; a small dot = non-prominent syllable. This transcription was originally invented to show intonation, but at this point we shall use it to indicate word stress. In (1.9), two words are transcribed with their citation pronunciation, which normally implics falling intonation — first with tonetic-stress marks and then in interlinear notation, e.g. (1.9) 'di:pa:t mentl 'di:kən tæmıneit 6. (=primary, = secondary, o = prominent, non-prominent). This notation, sometimes referred to as "tadpole", is a simplification of the interlinear notation; it accounts For all four types of syllable with no indication of pitch movement, c.g. brand-new caro differentiate In this work, we shall normally use the tonetic stress marking system, a sort of broad transcription which fulfills two roles: it displays accentuation by presence and pitch movement according to the shape of the mark; as explained before, prominence is left unmarked. The pictorial nature of the notation, which makes it relatively casy to interpret, together with the economy represented by the small number of marks it consists of, explain why the system has been successfully adopted in various intonation courses. The use of for primary stress shows the normal association of this stress with a falling nuclear tone, typical of citation forms. For instance, the citation pronunciation of the word departmental di:pa:t mentl/ contains all four levels of prominence/stress. If we number the syllables from 1 (left) to 4 (right) we find that syllable 1 bears secondary stress; syllable 2 is prominent; syllable 3 has primary stress, and syllable 4 is non-prominent. Notice, however, that in the citation pattern of the phrase departmental meeting /'di:pa:tmentl `mi:tın/ the syllable starting pitch contrast is no longer the third syllable, but the first. This phenomenon is explained in the next section. ### 1.7. Stress Shift By definition, a syllable bearing primary stress is the one which normally initiates pitch contrast; a secondary stress is less likely to be pitch contrast initiator, but it can certainly play that role when the rhythmic pressure of the context (i.e. the neighbouring prominences) causes the redistribution of stresses called ‘stress shift’. This change of roles is the result of the tendency in English to separate strong, primary stresses by weak, unstressed syllables. A similar tendency is not unknown in Spanish. The problem is that there is still a fairly amount of research to be done in order to identify the accentual behaviour of double-stressed items which reverse their stress patterns in a way similar to English when used in connected speech. This would explain why a word such as ‘veinti `cuatro, with two potential stresses, becomes 'veinticuatro `mil, and why the question ¿Estás de acuerdo? can either get the answer Abso 'luta`mente or Abso 'lutamente de a`cuerdo. This tendency towards stress shift scems more noticeable in English than in Spanish. The pressure exerted by adjacent stresses in Spanish, for instance, would make an English speaker say the names Iván Torres and 'Fidel Castro, rather than the normal Spanish patterns I 'ván Torres and Fi 'del Castro. Other patterns, of the BBC type /'bi: bi: `si:/, /'be be `se/- are the same in both languages, e.g. JVC, KLM, FBI, etc. Stress shift in English affects both simple and compound words whose citation pattern consists of a secondary stress followed by a primary stress in the same word. This double-stress pattern, however, accommodates to the pattern of stresses which precede and follow in the text, in such a way that stresses on consecutive (or almost consecutive) syllables are avoided. For instance, the accentuation of the phrase preconceived ideas comes from the citation patterns 'preconceived +ideas. The resulting pattern is, however, 'preconceived ideas, in which syllable pre-, originally carrying secondary stress, becomes the only pitch movement initiator in the word, and the original primary stress, -ceived, becomes a mere rhythmic beat for which there is no mark available in our system. (In other notation systems a mark is provided for these prominent syllables.) This stress pattern reflects the attributive function of the adjective, i.e. the adjective premodifies the head of the NP. In Your i 'deas are preconceived the adjective is predicative, i.e. it is part of the predicate, and consequently it is the secondary stress on pre- which is lost. Other examples with double-stressed simple words are, c.g. ``` (1.10) citation forms: attributive use: predicative use: 'atmospheric, 'audiovisual, 'auto`matic 'Let's di 'scuss 'atmospheric `pressure 'Why not 'use 'audiovisual `aids We 'made the 'trip on 'automatic `pilot The 'music was 'highly atmospheric The 'system we use is audio visual The 'process is 'fully auto`matic ``` Double-stressed compounds can also be subject to stress shift when used as adjectives or adverbs, either attributively or predicatively, c.g. ``` (1.11) citation form: 'double decker attributive use: 'Double-decker 'buses are 'typically `British predicative use: I 'love double-deckers (1.12) citation form: 'second-hand attributive use: I'm 'talking about 'second-hand `furniture predicative use: The 'books were 'bought second-hand (1.13) citation form: 'well-formed attributive use: 'This is an e 'xample of a 'well-formed sentence predicative use: 'Sentences 'must be well-formed ``` Similarly, proper names and names represented by initials can also undergo stress shift in the same conditions, e.g. ``` (1.14) citation form: 'Piccadilly attributive use: 'Let's go to 'Piccadilly Circus predicative use: She 'lives near Piccadilly (1.15) citation form: 1 'C'D attributive use: The 'new version's on 'CD `Rom predicative use: I have it on C`D ``` Stress shift also affects phrasal verbs, the citation patterns of which consist of two stresses, c.g. ``` (1.16) citation form: 'come out attributive use: The 'pictures 'came out 'well predicative use: The 'pictures 'didn't come out ``` A small group of single-, late-stressed adjectives, exhibit a similar tendency. Their predicative pattern is the same as their citation pattern, c.g. ``` (1.17) citation forms: attributive use: excess, com pact, direct We had to 'pay 'excess `baggage I'd 'like to 'buy her a 'compact `disc A 'transitive verb takes a 'direct object (1.18) predicative use: I'm a 'fraid he 'smokes to ex cess The 'new 'flat is 'very com pact We went to the 'beach direct ``` The above also occurs in a number of single-stressed place names, c.g. ``` (1.19) citation forms: Berlin, Belfast attributive use: 'When was the 'Berlin Wall knocked down? He 'lives on 'Belfast `Avenue predicative use: We went through 'East Berlin The 'capital of 'Northern Ireland's Bel fast ``` Finally, short phrases containing two stresses and which combine with further stressed material can also be affected by stress shift, c.g. ``` (1.20) 'very warm + `welcome = a very warm welcome (1.21) 'fairly `new + cre`ation = a 'fairly new cre `ation (1.22) 'good night + kiss = a good night `kiss (1.23) a 'hard `day + night = a 'hard day's `night (1.24) a lot more + rain = a lot more `rain (1.25) a 'lot `less + `money = a 'lot less `money (1.26) 'not very + much = 'not very much (1.27) 'much more + attention = 'much more at tention (1.28) 'far less + ex citing = 'far less ex citing (1.29) 'quite a bit + `older = 'quite a bit older (1.30) 'very little + time = 'very little `time ``` ## 2. Stress in English Simple Words ### 2.1 Introduction The first thing that needs to be clarified is to state what we understand by simple word. Strictly speaking, a simple word is one which is made up of a single grammatical unit called stem, e.g. photograph, judge. Derivatives such as photography, photographic and pre-judgement are not, technically speaking, simple words, but complex, because they contain affixes (either prefixes and/or suffixes). For EFL purposes, however, it will be better to consider all three examples as simple words. Stress in compounds will be dealt with in chapter 3. The second aspect is related to the way in which Spanish speakers can learn English stress patterns. The two main views on the subject are that: 1. The rules that govern English stress placement in words are so complicated, that students had better learn the pattern of cach word as they learn the word itself; 2. The learner should keep in mind the most common rules that predict stress placement in words and apply them whenever necessary; these are phonological rules, which establish relationships between phonemic patterning and stress (in §2.2 stress is related to English vowels), and morphological rules, which indicate the relationship between suffixes and stress (see §2.3) and also between grammatical class and stress, e.g. `progress (noun)-pro `gress (verb; scc §2.4). A further rule, which has to do with etymology, predicts stress placement according to whether the word is, for instance, of Germanic, Greek or Latin origin. However, given that our students of English are, on the whole, not familiar with diachronic studies, we shall not take up this analysis. Finally, a special difficulty is met with stress variation -the stress pattern of a few words depends on British vs. American preference and, furthermore, RP speakers sometimes vary in the stress pattern they use, the result being that a few words have alternative patterns (see §2.5 and §2.6, respectively.) Spanish speakers are recommended to make use of as many different types of information as possible in order to predict stress placement in simple words. They should also be aware of the large number of exceptions to the rules. Rules and main exceptions are analysed in the next sections. ### 2.2. Word Stress and Syllable Patterning The English phonological vowel system may be said to consist of two separate subsystems, strong and weak, cach one correlating with syllable stress. Thus, whereas stressed syllables can only have strong vowels, unstressed syllables can be made up of either strong or weak vowels. In the table below, the column ‘stressed’ indicates what vowel phonemes can get a stress in English syllables; the last two columns indicate what vowels make syllables cither ‘prominent’ or ‘non-prominent’. The following conclusions may be drawn from the table: 1. Any vowel apart from /ǝ, i, u/ -the so-called ‘unstressable vowels’- may occur with either a primary or secondary stress; /i, u/ are not to be confused with /i:, u:/, c.g. bikini /br`ki:ni/. The rest of the vowels may be stressed or unstressed. 2. All long vowels and certain short vowels always make syllables prominent; this means that they are inherently prominent. 3. /1, u/ belong to both subsystems. For instance, in the word minimum/mınıməm/, // plays the part of a strong vowel in syllable 1 and a weak vowel in the next syllable. Syllables containing /1, 0/ are prominent if they are stressed; this is because these vowels do not have inherent prominence. The last four types of syllables are always considered non-prominent. | | stressed | prominent | non-prominent | |---|---|---|---| | **Strong vowels (long): ** | | | | | i: | | | | | α: | | | | | u: | | | | | 3: | | | | | all diphthongs | | | | |