Is there a term for words which are pronounced differently depending on where they are in the sentence? For example, when I use the word "to" at the beginning or end of a sentence (or when I'm emphasizing it), I pronounce it the traditional way, i.e. the same as the number two. However, anywhere else, I pronounce it closer to "tuh" (i.e. rhyming with "duh").

Example of the first usage:

  • How many countries have you been to?

Example of the second usage:

  • Have you been to very many places?
  • 3
    It is not, exactly, where the word is in a sentence, but the stress one places on it. Although, yes, sentence location can be an indication of stress. So, terms to use include stressed or unstessed, or, speaking of the vowel, unreduced or reduced.
    – pazzo
    Sep 26, 2014 at 4:22
  • 1
    Almost every word will be pronounced differently depending on how it's used. Go look up prosody. Sep 26, 2014 at 8:00
  • 2
    @curiousdannii Yes, but not in that radical way. Go look up weak forms ;) Sep 26, 2014 at 10:13
  • 1
    @CarSmack but the examples of to in the OP's question are both unstressed!! Sep 26, 2014 at 10:14
  • The OP is just giving one type of example. They aren't asking specifically about only those. Sep 26, 2014 at 10:14

3 Answers 3


This is an interesting question. Several comments and one of the answers here make it sound like any word can go through this radical kind of transformation if it is used in a different part of the sentence. This is not true!!

We can split words up, generally speaking, into two types for our purposes here. There are the kinds of words that you look up in the dictionary to find their meaning. These are often called content words, or lexical words. I prefer the second term. These words normally carry the essential meaning of a sentence. So if a two-year-old kid walks up to you and says:


then you know pretty much what they're on about. Notice that these types of words are easy to look up in a dictionary and easy to explain. They consist of common Nouns, Verbs, Adjectives and Adverbs.

The second group of words are what people call function words, or grammatical words. These words are difficult to explain, and you wouldn't be able to understand them by looking them up in a dictionary. They're usually small and they're the kinds of words that you would study in a grammar lesson, especially if you're learning English. Importantly, they rarely take stress. They consist largely of Pronouns, Auxiliary Verbs, Prepositions, Determiners, Coordinating Conjunctions and Subordinators as well as strange words like infinitival to. These words are actually not very important or very helpful at all in terms of communication. Note that if a two-year-old walks up to you and says:

  • I TO TO THE!

.. you won't have the foggiest idea what they're trying to say! These words are the words that were missing, however, from the original sentence I want to go to the park.

Now the lexical words in a sentence usually contain the stressed syllables in that sentence. Stressed syllables in English give the impression of occurring at regular intervals. The effect of this is that stress and not the number of syllables is important for the perceived timing or duration of English sentences. The other effect is that the lexical words stand out more and the function words are less prominent.

In order to make this happen, English has a system for reducing the length of time it takes to say many of these function words that we've been talking about. This is so we can squash them in between the stresses. For many, if not most, of these words there are two forms (in terms of their sound). The usual form is the one used when:

  1. They aren't stressed
  2. They have another type of word or phrase following them, that they grammatically go together with.

These forms are called WEAK FORMS. They normally have a very short vowel. So the normal form for the Preposition to is the weak form /tə/ ['tuh' in the Original Poster's example]. However if to is used without a Noun directly following it it must be STRONG as in /tu:/. Note the vowel there has a colon-type mark. This indicates that the vowel is long. Also notice that the Noun is still there in the Original Poster's example, but it has moved to the beginning of the question and left to stranded at the end. We can draw up contrasts then such as:

  • Who are you looking /fɔ:/?
  • I'm looking /fə/ you!

Here we see strong and weak for respectively occurring without and with a following Noun. Note that strong for at the end of the sentence there is NOT stressed.

The rule is that prepositions like this are strong only if stressed OR if stranded. A strong form is definitely not an indicator of stress. We can have a full strong form vowel there without any stress at all.

As a rule of thumb then, if the word is the last word in the sentence it will be strong. If it isn't and it's not being stressed then it will be weak. [Pronouns are an exception though, they don't take another word and so they're usually always weak when not stressed].

One last point to make is that not all function words have weak forms. For example the words in and on don't have weak forms ( - perhaps we should say that they have no exceptional strong form - they're already quite short ...). These words are always the same.


To try and answer the Original Poster's question, the words whose pronunciation is most likely to be radically affected by their position in a sentence, are function words. Not all function words are affected by this process however. The different forms that are used are called weak and strong forms. Weak forms are the standard pronunciations and the strong forms are the exceptions. We should note that when we call a word by name, for example if we want to talk about the word to, we use the strong form. This is because a) we're stressing the word, and b) because we have dislocated the word from any following items.

Hope this is helpful!

  • In definitely has a weak form to me: the vowel (can) disappear(s) entirely and the nasal (can) become(s) syllabic. On only varies in vowel length (reduced extra short vs. unreduced normal short), which I'd agree is not really enough to call it a weak form. Oddly, to works just fine for me in its weak form when stranded, though for doesn't. Sep 28, 2014 at 21:40
  • @JanusBahsJacquet That's very unusual, in is generally reckoned to not have a weak form. We should get someone over to record you! ;) I've got a few friends with a nearpermanently weak to though... Sep 28, 2014 at 22:20
  • I suppose it might not really be a weak form after all, but just a regular vowel reduction, with the ‘extra unstressed’ reduced vowel being assimilated into the following nasal. I can't get myself to reduce similar vowels that much, though: “he’s in list #5” (leaving aside the awkward choice of preposition) has no vowel at all in rapid speech, while “he’s enlisted #5” has a schwa that's not assimilated to the nasal. Sep 28, 2014 at 22:26

I am unaware of a single word that encapsulates the idea, but the topic usually comes under the phrase word stress in English. This is commonplace when it comes to words that are used as both verbs and nouns.

They set a world record.
They're recording the song this fall.

The pronunciation (i.e. syllable stress) is different. Where "record" is used as a noun, the stress is on the first syllable: RE-cord (where "re" is the same sound as the "re" in "relative").

But when "record" is used as a verb, the stress is on the second syllable: re-CORD, and the "re" sounds like "ri", as in "remember".)

Some other examples would be:

  • Increase
  • Decrease
  • Import
  • Discount
  • Refund
  • Permit
  • Conflict
  • And some others: police, cement Sep 26, 2014 at 13:19
  • @guifa They're both used as verb and nouns, but the stress is on the same syllable for each: po'lice, ce'ment. Unless it's different across the pond? Sep 27, 2014 at 8:33
  • 2
    @Araucaria for me (SAE) it's "the POlice poLICE" and "one ceMENTS with CEment" Sep 27, 2014 at 13:32
  • @guifa Cool. That's really interesting. Thanks. Wonder why the RP doesn't follow the regular pattern ... Sep 27, 2014 at 13:44
  • @Araucaria Would you folks tell me what SAE and RP are? Sep 28, 2014 at 9:02

To add to CarSmack's comment, you may want to look into the term, intonation, for the apparent difference in "pronunciation" although it is the pitch (high/low) and length of vowels that provide the word with a different "tone."

In theory, any word (with a vowel, of course) can have such effect you described.

  • You can't really do this with any word ... You can say I kn dance for example, but you can't say The mn you saw yeasterday ... Sep 26, 2014 at 10:37
  • @Araucaria except your depicted words do NOT have a vowel. I do your see somewhat manufactured point; after all, the OP tagged this as "pronunciation." Sep 27, 2014 at 13:49
  • Sorry, crosscounter not quite sure which depicted words you mean. Do you mean can /kn/ ? Or for? Sep 27, 2014 at 13:52

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.