Let's say an IPA pronunciation contains a double consonant, such as "dd" or "ss". Does that really mean this consonant should be pronounced twice?

There are examples where this would indeed be the case, such as the German word "arttypisch" /ˈaːɐ̯tˌtyːpɪʃ/ that requires two consecutive stops of the tongue for the two "t" sounds. (In this case, it's because "arttypisch" is a compound word make up from "art" and "typisch" and because pronouncing it with just one "t" would make it sound like the different word "atypisch".)

But browsing Wiktionary, I've come across many pronunciations where I suspect that speaking both consonants separately would be highly uncommon, if not outright wrong.

So I wonder:

  1. Is there some rule as to when a double consonant in an IPA pronunciation should be spoken twice, without knowing about the word's etymology?
  2. Am I right in assuming that across languages, the vast majority of cases requires the consonant to be spoken just once?

Here are some more examples:

Language Word Pronunciations Source
English abstinency /æb.stə.nn̩.si/ en.wiktionary.org
English excentric /ɛksˈsɛn.tɹɪk/ en.wiktionary.org
German Brieffreund /ˈbʀiːfˌfʀɔɪ̯nt/ en.wiktionary.org
German deutschstämmig /ˈdɔɪ̯t͡ʃˌʃtɛmɪç/ en.wiktionary.org
Italian allegretto /al.leɡˈret.to/ en.wiktionary.org
Italian capello /kaˈpello/, /kaˈpɛllo/ en.wiktionary.org
Italian correggere /korˈrɛddʒere/ en.wiktionary.org
Italian ufficioso /uf.fiˈtʃo.so/, /uf.fiˈtʃo.zo/ en.wiktionary.org
Italian terranoce /ter.raˈno.t͡ʃe/ en.wiktionary.org

Short answer: yes, it generally means the same consonant twice, but that doesn't necessarily mean there's a gap in between them.

If you're a native English-speaker, think about how you'd say "acting" or "lapdog" in normal conversation. For me, unless I'm deliberately trying to enunciate, there won't be any release between the two stops. Now think about "bad dog" or "black cat" (again, in casual conversation, not specifically enunciating): again, for me, there's no release in the middle, just a single longer stop.

This is what a doubled consonant usually means in IPA transcription: a consonant pronounced for twice as long, but not necessarily having any additional release in the middle. This meaning is fairly similar to ː (the length symbol), but has different implications: the double N in Latin /annʊs/ "year", for example, acts like a sequence of two consonants for the purposes of syllabification, while */anːʊs/ would imply it's a single phonemic unit. Something similar happens in Italian: look at where the syllable breaks fall in /al.leɡ'ret.to/.

  • 2
    "ː ... would imply it's a single phonemic unit." Not necessarily. The length mark is often employed for Swedish and Norwegian but they are not distinct phonemes. It's just conventional as far as I can tell. – Nardog Feb 12 at 3:59
  • 2
    If you’re using strictly phonemic notation, then yes, /Cː/ should (ideally) indicate a phonemically long consonant, whereas /CC/ indicates two sequential identical phonemes. In phonetic notation, conversely, /Cː/ indicates that the sound consists of one consonant held for a longer period of time, whereas (in theory, at least) /CC/ indicates two fully and independently produced consonants. So arttypisch is not normally [ˈaɐ̯tˌtyːpɪʃ] in phonetic terms, since that entails separate releases of the two t’s. It can be written [ˈaɐ̯ˌtːʰyːpʰɪʃ] or (more intuitively) [ˈaɐ̯t̚.ˌtʰyːpʰɪʃ] instead. – Janus Bahs Jacquet Feb 12 at 22:12
  • … Whereas ‘goodbye’ phonemically would be /ɡʊdˈbaɪ/, with two separate phonemes, but phonetically [ɡʊˈbːaɪ], with what is acoustically a single, long consonant (or [ɡʊb̚.ˈbaɪ], if you prefer). – Janus Bahs Jacquet Feb 12 at 22:16
  • +1 Technically (aka 'pedantically'), in a word like acting, there will be a release of the /k/, but it will be 'masked'. In other words the air from the /k/ will be usually be released after the closure for the /t/ is made, and will thus usually be be inaudible. However, it is completely necessary, because otherwise there would be no build up of air pressure behind the alveolar stricture and no [t] would be heard! – Araucaria - Not here any more. Feb 13 at 22:50

It does not mean that. In fact it has no intrinsic meaning. It is possible that some language has a contrast between consonant clusters with distinct articulations versus single consonants with longer duration, in which case one could use the C: vs CC notation to indicate that. You would have to read the discussion on whatever article you find this usage to see what if anything the author says. On phonological grounds, geminate consonants act like clusters of consonants (with a few special properties) so when people insert syllable breaks they often write [t.t] to indicate that the consonant is both in the preceding and the following syllables. Some language don't have "geminate consonants" as a special kind of consonant (geminates are phonologically distinct in Italian, Finnish, Japanese, Luganda, Arabic, Icelandic) so in English which does not have geminate comsonants, surface "dd" as in "sad dog" is simply one d followed by another. The transcription /ɛksˈsɛn.tɹɪk/ is dubious; but {æb.stə.nn̩.si] is not dubious, though it comes from æb.stə.nən.si where two two n's are not next to each other. This does point to common ambiguity in the use of "consonant", where sometimes it means "C", "not a syllable peak" ([n̩] is functionally a vowel) and sometimes it means "non-vocoid", something with more constriction.

  • Thanks for pointing me toward "geminate consonants"! This helped me find further relevant material. – Daniel Wolf Feb 13 at 13:58

For the examples you listed, you indeed pronounce both consonants and this is not uncommon. They are still used like this up to this day. Just check the Oxford and Cambridge dictionary. You may not be familiar with this way of pronouncing but this changes nothing to the fact that the correct pronunciation is by pronouncing both consonants. In normal conversation and when we combine words especially in English we tend to use weak forms of those consonants and blend them with what follows or what comes before. A good example for this is the word spaceship. When putting space /speɪs/ and ship /ʃɪp/ together we get spaceship /ˈspeɪs.ʃɪp/. But when we pronounce the word we normally do not pronounce those words as being separated but rather tend to combine them and also blend the phonemes connecting those two words.

You can't really put German and Italian on the same step as English tho, because especially German is a language that works completely different from English. In German we tend to make breaks after each word and part of speech. In English we try to combine everything. If you look for example at sound files you will see that recorded speech in German will always have breaks in between words and sentences, whereas English won't feature this.

Coming to your two questions.

  1. There is no real rule. You have to know the statistics and act according to this. It is the same just like with collocations. You need to get a feeling for this and then you start to use it.

  2. This is not true. The way the IPA transcription is written is also the way how the word is to be pronounced in its standard strong form. Where in English you have the possibility to blend those consonants sometimes, in German this would be a mess that changes the meaning of most of the words. Also for languages like Spanish and Portuguese your assumption is incorrect. All those languages have specific rules how to pronounce those special cases. Sometimes a double consonant causes into having a long consonant, sometimes short and sometimes a double one and so on.

  • If you are interested in this field of study I could look in my older databases and give you some lectures and papers that exactly deal with this topic. – Jannik Feb 12 at 18:38
  • 1
    -1 “In German we tend to make breaks after each word and part of speech.” <— Not at all true! – Araucaria - Not here any more. Feb 13 at 16:27
  • @Araucaria If they're talking about glottal stop insertion, it seems correct to me. – Nardog Feb 14 at 14:00
  • @Nardog But glottal stop insertion occurs before word-initial vowels. OP is discussing CC sequences! Also, glottal insertion, or hard attack, also occurs in UK English and even more frequently in US English. And even then that hardly amounts to the claim that German speakers "tend to make breaks after each word and part of speech"! (What's the difference between a word and a part of speech?) As for "If you look for example at sound files you will see that recorded speech in German will always have breaks in between words and sentences, whereas English won't feature this" The mind boggles! – Araucaria - Not here any more. Feb 14 at 16:47
  • @Araucaria-Nothereanymore.We do make stops after each word. I am a German speaker... There are several studies and books talking about this. For example Longman Grammar. If you look closely into the audio files and you know how to analyse them correctly you will see this as well. Where in English we tend to combine sounds with each other and blend them, in German we make stops most of the time. We do not have this blending like many other languages. – Jannik Feb 15 at 18:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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