Are sentence stress and word stress analyzed seperately? Or are they analyzed all at once? For example "I need to sell my car". Do I find the sentence stress and word stresses seperately or do I analyze them all at once? If I analyze them all at once, Does one affect the other? Can sentence stress turn a secondary stressed syllable in to a primary stressed syllable? Or can it turn a non stressed syllable in to a primary or secondary stressed syllable through sentence stress? I'm asking these questions only referring to English.

  • I am not a trained linguist, but understand "word stress" to mean what syllable is stressed in a word. Your example only includes monosyllables and so word stress is not an issue. I understand "sentence stress" to mean "prosodic stress as a part of intonation. If you want to combines the two in your question, you may need an example that includes multisyllabic words. Commented Jul 5, 2022 at 20:48
  • @Vegawatcher its not that I want to combine the two, its that I wonder whether the two are analyzed toguether or not. How about the following example: I wanted to sell my car.
    – Luiyo
    Commented Jul 6, 2022 at 0:05
  • A few final questions for clarification, since your question has potentially a huge scope. By analysis, are you asking (1) about how to predict the surface stress accents taking into account both word and sentence stress, (2) about how a linguist would analyze the surface stress phonetically, or (3) about how a linguist would analyze the surface stress phonologically? The first focuses on the rules; the second focuses on how the stresses sound; and the third focuses on what the stresses do in the language. Commented Jul 6, 2022 at 15:10
  • @Vegawatcher I'd say both 1 and 3.
    – Luiyo
    Commented Jul 7, 2022 at 1:06

2 Answers 2


As far as I am aware, English has phonemic stress on at least three levels: word stress, compound stress, and sentence stress. In all three cases, the stress manifests as a differentiated syllable or two within a single word. To analyze why a syllable is stressed in a stream of words, you have to examine all three levels.

English is a stress-timed language, which means that normal speech rhythm tries to space each stress with the same time interval, while adjusting the rate at which unstressed syllables are said to accomplish this.

The stress functions phonemically to give prominence to one or more syllables within a word and typically contains multiple phonetic components, including an increase in the syllable's loudness, a higher pitch, full vowel articulation, length, and perhaps a reallocation of the phonological feet within a word.

Every word spoken in isolation, even if monosyllabic, must have at least one stress. In fact, the only utterance I can think of that has no stress is the hesitation sound--often spelled "uh"--,because it is meant to signal something else to be said that would carry a stress.

Some of the components of stress may be deleted because of other phonological processes, but some type of syllabic prominence is normally retained even in "extreme" phonological environments, such as singing or utterances attempting to contrast unstressed syllables in similar words. For example, with yes-no question intonation, a sentence generally has rising intonation, therefore causing a final unstressed syllable to be articulated at a higher pitch than previous stressed ones. Such a change does not affect the perception of what syllable is stressed, because of the other feature of stress that will remain.

For word stress, I believe linguists generally recognize up to three potential levels that can occur in one English word: one syllable with primary stress, one syllable with secondary stress, and an unlimited number of unstressed syllables. Every word must have at least one stress; however, normally stressed syllables are routinely de-stressed by other processes, frequently leaving words without a stressed syllable when not used in isolation.

Word stress often causes vowel reduction and centralization. There are certain vowels that are exempt, and others that can vary from speaker to speaker or even utterance to utterance.

The word stress is a suprasegmental feature that helps differentiate words. Over at least the last two millennia, there has been a strong tendency to move the stress to the first syllable, especially in nouns, causing many minimial pairs in which a noun is stressed on the first syllable, but a corresponding verb is not--e.g, " a setup" vs "set up something"; "progress report" vs. "progress with the report"; and "a record performance" vs. "record a performance."

Compound stress takes word stress as an input and creates a new phonological word. English has at least two patterns of intonation used in compounds, one of which leaves the stress on the modifier and removes the stress on the modified part. Examples are "townhouse," "dog house," and "White House," in all of which the component "house" is not stressed. This pattern contrasts with the stresses in "town hall," "a dog's house," and "a white house," in which hall/house would usually retain a stress.

The type of compound stress used in "townhouse" differentiates this type of compounds from other compounds. I am unsure of the precise semantic value, but I think it creates a subcategory of the modified noun, whereas the other stress pattern (i.e., the one used in "town hall") seems to have the semantics of a reduced noun phrase without creating a new subcategory.

Sentence stress takes word stress and compound stress as an input and uses this stress to signal information value. The stress is removed from any element that is already part of the discourse, easily inferable, or otherwise of low information value. New or salient information retains stress or even receives additional stress. This process greatly affects the rhythm of the syllables to allow roughly equal spacing between each stressed syllable.

The fact that sentence stress signals information value explains why it can cancel the other functions of stress to differentiate words and compound types. The lack of stress indicates that these differentiating functions are no longer needed in the specific utterance.

The operation to remove stress is part of the pragmatics of an utterance. Even though some aspect of it is mandatory in normal speech, the pragmatics are mostly under the spontaneous control of the speaker. The tension between what is mandatory and what is under spontaneous control of the speaker emerges when reading the words of others, especially poetry, where the reader's spontaneous assessment of the pragmatics may differ from the assessment of the original author and result in mistakenly reading in a stress pattern that clashes with the pragmatics of words later in the speech stream or with the meter of poetry. To avoid such clashes, good readers must read with attention to the overall pragmatic flow of an utterance to anticipate what is needed or even prepare a reading in advance so as not to be surprised by an unexpected intonation or rhythm.

A reader that produces a phrase with malformed stresses will often feel obligated to reread the mistaken part to produce an acceptable utterance. Good prose writing tends to avoid relying on stress patterns not easily predictable from the speech stream.


English stress is analyzed extensively in The sound pattern of English, and a number of rules are required to generate the patterns which they observe (a system of numeric stress degrees). “Stress” is treated as a unified phonetically-defined thing, thus it makes sense to talk about “word stress” and “sentence stress”. There are many rules involved in generating word-internal patterns versus phrasak patterns. It is a basic architectural feature of the system (due to cyclicity) that first you generate word stresses, and that output determines how sentence stresses are computed.

Subsequent developments in the theory, for a while, changed the representational basis for “stress” but maintained the view that word stress and phrasal stress are the same kind of think. However, over the past few decades the dominant view is that word-level stress representation is much simpler: stressed vs. non-stressed, with the option of a distinction “primary” vs. “secondary”. Then, at the level of phonetic implementation, particular tonal objects are aligned to prominent syllables within words – the “stuff” that constitutes sentence-level stress (tone melodies of sort) is entirely different from word-level stress (prominence). Prominence within the word comes in two flavors. First (topmost) is foot prominence – syllables are grouped together into rhythmic groups of two or three syllables, and some foot is the prominent one within the word. Then within the foot, one syllable is marked as the prominent one.

In something like "I need to sell my car", every word is monosyllabic, so word-internal stress is irrelevant, you just have to analyze the pitch patterns at the sentence level. This introduces one of the main complications of phrasal "stress" i.e. intonation: any of those words can be focused (except "to"). Whatever the general tone pattern of the utterance is, its realization is partially dependent on whether some word is focused, hence "I need to sell my car; I need to sell my car...".

There is a (former?) claim that stressed shift around inside the word, at the sentence level. Example: "sixteen; sixteen children". This phenomenon of "rhythm" has been analyzed as a transfer of stress within a word from syllable-to-syllable; but it has also been analyzed as the suppression of prominence in certain phrasal positions. See this paper for some discussion. IMO the argument for phonological sentential stress-swapping is weak.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.