In Arabic it's an ordinary phoneme, and hence, a letter in the alphabet, the hamza.

In French it's one of the ways not to pronounce an h, as in hôtel.

In many regional variants of English it's used to pronounce the t: le'er for letter, etc. It's not a separate phoneme.

In uh-oh (which I think in Dutch is a loan word from English) it is definitely essential to its meaning, but does one word a phoneme make?

Strictly phonologically speaking, most English and other Germanic words we think of as beginning with a vowel or silent h actually begin with a glottal stop when spoken individually. Uh-oh, to take an example from a previous writeup, contains two glottal stops, one before each vowel. Feel your voice box when you speak it if you don't believe me. Unless you're a native speaker of a language that doesn't regularly use glottal stops, the two vowel sounds are almost definitely separated by a voiceless consonant articulated by closing your glottis.

Glottal stops also appear more frequently in many English dialects (try a Scottish "accent" on for size) where more "standard" dialects use alveolar stops like d, t, or flaps.

The glottal stop is a sound used in speech in many languages. It is a consonant made by momentary closing the glottis. As it is a momentary pause in the flow of sound, it is usually written using the apostrophe ( ' ).

The glottal stop is not part of the English language as it is taught and spoken in most of the world 1. However it is characteristic of the way that many of The English have come to speak English from the late 20th century onwards.

Many English people will soften or drop entirely the t at the end of words such as that, not, but, what, got. I tend to write this accent as wo' for what and so on.

The glottal stop is even more emphasised in some accents. Instead of t and th in the middle or at the end of a word. For instance, an east-end Londoner may say:

Wi' a li'il bi' of bu'er, no' a glo'al stop. ge' i'?

Here endith the objective, factual writeup. Now for opinion and GTKY stuff.

Note 1) That used to read: "The glottal stop is not officially part of the English language. However..." until a britnoder asked me to define official in this context. I have caved in and admitted that in these postmodern, decentered times that is not possible. Everyone has an accent. All an official pronounciation would be doing is arbitrarily promoting one accent over another, and that would be bad m'kay?

As Princess LouLou keeps saying to me, English is a living language. It evolves over time. There is no single right way to speak, no official correct pronounciation.

She also tells me that my accent has softened over the last few months since I arrived in London. This is good. Despite that, I try not to use glottal stops in my speech. Many of the britnoders do, and as I assimilate the local accent I'm trying to avoid doing so. I've gotten used to hearing it, I just don't want to hear myself say it. It just sounds ugly. You can take that as a value judgement if you so wish.

The term 'glottal stop' is perhaps the only one from phonetics that has entered common use. It is a normal sound in many languages, though has not yet become one in standard English. I shall first give examples from other language groups, then discuss the phonetic articulation in more detail. The SAMPA symbol for this sound is [?].


It is common in the Semitic languages. It is the original value of the Hebrew aleph and the Arabic letter alif, the first of their alphabets. Since the glottal stop did not occur in Ancient Greek this letter was adopted as alpha for the vowel A.

I don't know enough about Modern Hebrew to comment, so the following is solely about the Biblical language; and the Arabic is Quranic.

It can occur initially and medially in Hebrew words. In some words aleph was not pronounced: לא lo 'not', ראש rosh 'head'. In these cases it is after a vowel but not followed by another.

In Arabic it can occur in any position, as in رأس ra?s 'head'. In the other corresponding word it has become silent: لا laa 'no'. The letter alif no longer represents the glottal stop, but has become a bearer of vowel signs and the marker of the long vowel aa. When the glottal stop is actually pronounced it is marked by an additional hook-shaped symbol ء hamza, usually on top of the alif as أ but occasionally elsewhere, depending on the phonetic context. The name hamza is sometimes used generally in linguistics for a glottal stop.

In the Semitic languages, words could not begin with a vowel. Any word that we transcribe with one actually begins with either the glottal stop or another sound ayin/ain, a pharyngeal fricative.

However, in Maltese, a descendant of Arabic, the original glottal stop has been lost entirely, but the Arabic sound q has changed into it, and the Maltese letter Q represents a glottal stop.


The Austronesian family covers Philippine, Indonesian, and Polynesian languages. In Philippine languages such as Tagalog, typically the glottal stop is used as a separator between vowels such as aa or oo, instead of treating them as long vowels.

In Indonesian, a word can't begin with a vowel. It can end with a vowel, but can't end with k. In the word kanak 'small child' the final k is a glottal stop. In the word anak 'child', there is an unwritten glottal stop in front of the first vowel. It does not really occur between vowels, though it may be heard in the odd word like maaf 'excuse me'.

In many Polynesian languages an original consonant k has changed into a glottal stop. For example, the legendary homeland of the Maoris is called Hawaiki, corresponding to the Hawaiian name Hawai'i and to the main Samoan island, Savai'i. (In Hawaiian a new k sound has arisen from earlier t: for example Maori tiki is Hawaiian ki'i.) The Hawaiian name for the vertical stroke symbolizing it is 'okina. The sound occurs between vowels and initially. No consonant can occur at the end of a syllable in Polynesian. In Samoan the consonant is described as "weak", a bare hiatus between vowels, and serving to shorten a vowel initially.


The sociological situation with English glottal stop has been discussed in previous write-ups. Phonetically, there is not just one simple replacement going on. This is one of the gradual changes I discuss in Changes in Southern British pronunciation: for most people, even fairly conservative speakers, they probably use glottal stops before another consonant, as in not really, in casual speech. Different speech registers have different phonetics even in the one person. The marked feature of the London Cockney accent and the more widespread modern accent called Estuary English is that these stops are used in a wider range of positions, the process known as glottalling.

Even in more conservative Received Pronunciation, where the t is not replaced entirely, a phenomenon called preglottalization can occur. I am not clear about this and don't know how much it occurs in my own speech; I would have to read more about it and test myself. But briefly, some stops can be preceded by a weak glottal stop in some positions: in ma'ttress but not in hopeless.

It is sometimes heard between words to prevent a linking R: as in India Office said as India'Office rather than IndiarOffice.

The glottal stop can occur initially in English in emphatic expressions like uh-oh and absolutely awful!, but normally it does not.

In German and Dutch however, words never begin with a vowel, except when closely connected with a preceding word. Normally a word written with initial vowel has a preceding glottal stop, as eins 'one' = [?aIns]. It remains when a prefix is added, as in geantwortet 'answered' from antworten.

In Danish there is the stød, not a distinctive glottal stop phoneme in its own right, but a simultaneous glottal break that can occur on both consonants and vowels.


In some respects it is a stop like any other: a brief closure of the mouth or throat, with abrupt release. But since the closure is made in the glottis (or larynx), the source of many features of speech, such as voice, it is unusual. It cannot be voiced, for example. While a bilabial stop [p] is blocked at the lips, it can simultaneously admit continued vibration through the glottis, giving the voiced counterpart [b]. This coarticulation between the bilabial and glottal positions is not possible for the glottal stop itself.

The preglottalization of RP English and the stød of Danish are examples of coarticulation of a glottal stoppage with some other sound. In some of the tonal languages of South-East Asia, such as Burmese and Vietnamese, one of the tones is "creaky", that is has a catch in it. When the vocal chords in the glottis are vibrated so as to give the effect of repeated constrictions through a sound, this is called creaky voice, though this is a kind of continuing voice, not a complete stop.

The most common way a glottal stop is coarticulated with another sound is in the class of stops called ejectives, occurring in Georgian, Amharic, Quechua, and many other languages. In these there are not just two places of stoppage, one of which is the glottis, but the glottis is moved upward to compress the trapped air, so that the release has a more forceful pop.

This, I believe, is the reason why we say a few languages, such as Arabic, German, and Indonesian, obligatorily have glottal stops before their initial vowels, whereas most languages don't. In one sense, the opening of the vocal chords when you begin to speak is a kind of release of a glottal stoppage. The reason we don't hear it as a glottal stop in English is that it's a normal plosive release, of no great strength, whereas the hamza-initial languages like German and Arabic have a glottal ejective in this position.

This might also explain the nature of the weak Samoan stop, as in fa'a-Samoa 'in the Samoan manner', described as a mere hiatus. It might be that this is a glottal plosive, whereas in London English [fla?a] 'flutter' or Arabic [bada?a] 'he began' it's a glottal ejective. But this bit is my own speculation.

Log in or registerto write something here or to contact authors.