I am not understanding how a word can begin with a glottal stop? Is it a glottal plosive? I guess I am trying not to outright ask why is it not called a glottal plosive.

When I say some words that are supposed to begin with a glottal stop, there is a slight puff of air, but not aspirated, and when I hear words, like the Arabic words for mother, I just hear a vowel.


4 Answers 4


English words with vowel initial tend to get a glottal stop. This occurs in most dialects, so a native speaker wouldn’t notice its presence or absence; they will just hear it as a “normal” vowel.

Formally speaking, a glottal + vowel is perceived as an allophone of that vowel alone.

However, there is a significant difference between pronunciation of word-initial vowels at the beginning of an utterance and in its middle. Words that begin with vowels get a preceding glottal stop only if they are at the start of an utterance. Compare apple and the apple (the rhymes with bee). There will be no glottal stop in the second case, and it is very easy to notice.

In other languages (like Hawa'iian), the presence or absence of the glottal stop (including the initial position) changes the meaning. One example is ʻai (with the initial glottal stop). It means food, and one should not use its counterpart word with a plain vowel at the beginning.

  • Thank you, I see. So I can make myself say the apple with a glottal stop. I see. Commented Nov 21, 2017 at 5:28
  • 1
    As a minor addition: There are also languages like German, where words beginning with a vowel always get a glottal stop, even in the middle of a sentence and even though the glottal stop is not phonemic. Commented Nov 21, 2017 at 10:35
  • Is there a reason why you did not analyse these (orthographically vowel-initial) words like angel or ant or opthalmology as beginning with a null phoneme? And glottal stop as an allophonic variant of the null phoneme? This explanation fits with things like spoonerisms. Commented Nov 21, 2017 at 15:57
  • 1
    @Героямслава Is there a reason why a null phoneme should be transcribed? If it's null, it's not there; why transcribe it? Transcription is for things that are there.
    – jlawler
    Commented Mar 20, 2023 at 16:38
  • 1
    @ErikE I bet you do know how, and do it unconsciously every day of your life. If you're an English speaker, it's more likely that you don't know how to NOT begin a vowel-initial word with a glottal stop at the beginning of an utterance.
    – Graham H.
    Commented Mar 21, 2023 at 13:35

In phonetics, there is a distinction among:

hard vocal onset - effectively this is most typically the glottal stop; the exact realisations sometimes differ (e.g. strident phonation / harsh voice instread of a stop) but that corresponds in the variation of of glottal stop phonemes too; hard voice onset is typical e.g. for Czech, where most of the times all word initial vowels start with glottal stop.

soft vocal onset - this happens when the glottis does not get so tense and the vowel starts pretty much right away; this is typical for French (you can have hard voice onset there too, but it carries certain markedness).

aspirated vocal onset - again glottis is not fully tense and the vowel starts with brief [h]-like sound; from the top of my head I do not recall any language using this, but if I remember correctly it may be found in some English dialects.


There are two distinction questions here: is it physically possible, and is it linguistically significant. It is clearly physically possible to have an initial vowel without a glottal stop, also with. What is much less clear is whether there can be a surface contrast between a bare utterance-initial vowel and an utterance-initial glottal stop plus vowel sequence. That is, can [ama] contrast with [ʔama]? Areare is a languages reputed to have a contrast between initial vowel versus initial ʔV; to detect the difference, you have to put a vowel before the word, in which case it is no longer utterance-initial. I cannot say that there is actually a phonetic distinction made in that language (I only worked on it for a quarter). So there can be no doubt that glottal stop can contrast with Ø (it is a real phoneme), I do not know of a phonetic study that demonstrates an utterance-initial distinction, as exists in the case of "pit" versus "bit".

  • When I say ama I pronounce it [ama] but not [ʔama]. I am pleased, after saying that, I notice I am not immediately pronouncing the american pronunciation of apple as [ˈʔæpl̪̩ˠ] but as [ˈæpl̪̩ˠ]. Commented Nov 25, 2017 at 0:45

When you try to remove word-initial glottal stop [ʔ] before vowels, it turns to voiced glottal fricative [ɦ].

First of all, [Ø] (no consonant) is not what you think it is. It exists as a phonemic element (/Ø/), but phonetic initial vowel [ØV], a bare utterance-initial vowel, doesn't actually exist. I'll use [] for phonetic means and // for phonemic in the rest of this post.

When we make a sound for speaking, it requires some air flowing through our glottises without exception. When trying to remove the initial [ʔ], due to the airflow, it becomes [ɦ] (voiced glottal fricative) instead of "[Ø]" which we desire. The characteristic of [ɦV] is that the consonant completely merges with the vowel, unlike ALL other consonants. The ways to pronounce [ɦV] and [ØV] are generally the same: form the vowel, then let the air flow to vibrate the cord while all other parts of our vocal tracts are at their default (relaxed) positions. The only difference is between vowels, where [ɦ] aspires more but [Ø] stays the same. For example, Japanese word 情報 [dʑoːɦoː] (information) has extra aspiration at [ɦ] (/h/) than 女王 [dʑo'oː] (queen).

It won't be a problem when [ɦ] and [Ø] are between vowels because the air is already flowing. However, it will be a dilemma to distinguish initial [ɦ] and [Ø] because the air is just about to flow, thus no comparison of "extra air" right after it breaks silence, or we can say it is always extra air comparing to no air. If you think you can get initial [ØV] by pushing less air, you are wrong: it will be just a quieter [ɦV], and a consonant cannot be determined by its volume.

Another fact to see is that the ancient alphabet sets, like Proto-Canaanite and Phoenician, have glottal stop (the origin of alphabet "A") as their very first alphabet. The past people already recognized glottal stop as "default" of a syllable, that's why many people didn't realize the existence of initial [ʔ] in their languages until modern days.

So, you won't find any difficulties to distinguish initial /ʔ/ and /Ø/ even if your native language doesn't have phonemic differences between [ʔ] and [Ø]. The problem is our vocal system disallows removing both [ʔ] and [ɦ] before an initial vowel. Luckily, Hawaiian language has /Ø/, /ʔ/, and /h/ but not /ɦ/, thus no problem to recognize initial [ɦ] as /Ø/ and intermediate [ɦ] as /h/.

There is no way to avoid [ʔ] becoming [ɦ] in your attempt, as the "hardware", our vocal system, just doesn't support pronouncing without air passing through the glottis.

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.