11

I've been learning Ancient Greek for about a year. The consensus amongst scholars appears to be pretty unequivocally that "p", "t" and "k" sounds had aspirated and unaspirated versions of an underlying consonant which were different phonemes, e.g. φ/π (whereas in Modern Greek these aspirated versions became fricatives a long time ago).

Obviously there are Hindi etc. speakers in Europe, but are there any existing indigenous European languages where aspirated/unaspirated versions of consonants which are different phonemes? For linguists I'm not sure where the boundaries of Europe are said to be, conventionally: I presume Turkey for example is not considered to be part of Europe, but that Georgia is, but in fact I have no idea what the thinking about that point is.

For (belated) clarification: I am looking for languages where native speakers and native listeners MUST be able to distinguish, in speech, between an aspirated consonant and its unaspirated counterpart, failing which mistakes in meaning will be made. Native speakers of Ancient Greek would fail to communicate properly if they could not master the differentiation while listening or speaking.

In the case of Ancient Greek, all 3 consonants concerned are voiceless. In the case of voiced consonants, it would be necessary to find a language where you absolutely MUST be able to distinguish between, for example, voiced unaspirated [b] and its voiced aspirated counterpart [bʰ], or risk not being able to communicate. You also absolutely MUST be able to make that differentiation when producing the sounds (speaking).

10
  • 1
    φ & π are very definitely distinct phonemes, and not allophones (likewise τ & θ and κ & χ). Can you clarify your question in that light? Commented Nov 20, 2024 at 10:23
  • So would these be "phoneme pairings"? Is that the right term... will edit. Because obviously these are 2 versions (in each case) of an underlying something... Sorry, struggling with the terminology here... Commented Nov 20, 2024 at 10:24
  • it's a phonemic contrast between voiceless aspirated and plain voiceless stops. It's worth noting that this is no more (or less) two versions of the same thing than between voiceless and voiced stops. Both are contrasts of voice onset time so rather than a pair of phonemes π & φ, you have a triad of phonemes β, π, φ (likewise δ, τ, θ and γ, κ, χ) Commented Nov 20, 2024 at 10:29
  • ftr, allophones are positional variants of the same phoneme (or meaningfully contrasted sound). They are different ways that the phoneme can be pronounced depending on its environment. For example, Spanish has a phoneme /d/ which is pronounced [d] after a pause, /n/, or /l/, but which is pronounced as [ð] (roughly the voiced th sound found at the beginning of the English words "this" and "that") in other positions. In this case, [d] and [ð] are allophones of the phoneme /d/ Commented Nov 20, 2024 at 10:43
  • 1
    I think you are asking about minimal pairs, not allophones. Commented Nov 20, 2024 at 11:06

4 Answers 4

16

Well, for all reasons Romani counts as an indigenous language of Europe. From the fourfold distinction in most Indoarian languages on stops Romani went to a threefold distinction: voiced–voiceless–voiceless aspirated.

So one finds the segments b, p, and pʰ; d, t, and tʰ (only one kind of this, the retroflex stops were rhotacised in Romani); and g, k, and kʰ.

7
  • 1
    Thanks. As may be apparent by now, I'm not a linguistics expert: when you say "one finds the segments" ... does this mean that if you fail to master the differentiation p/pʰ t/tʰ or k/kʰ ... you risk getting the word wrong and inadvertently saying "kettle" when you mean "shirt"? I.e. are these contrasts actual determiners of different meanings in Romani? Commented Nov 20, 2024 at 14:46
  • 1
    Yes. That's what the segments (more precisely, phonemes) are for. Distinguishing meaning. Commented Nov 20, 2024 at 15:05
  • Ah, Romani – good spot, didn’t think of that! Commented Nov 20, 2024 at 16:31
  • I wonder if there are others? Janus's answer seems (possibly) to claim that English is in this category. If so, I don't agree (bearing in mind the attempt at clarification I've just made in the question). I'm now going to try to seek out some Romani, to see how this differentiation is done, and how hard it is for a non-Romani speaker to get the differentiation in practice. Commented Nov 20, 2024 at 20:14
  • 1
    @Sixtyfive It’s not that Romani innovated /pʰ tʰ kʰ/, but that the ancestral system with four distinct types (/b bʰ p pʰ/) was reduced in Romani to one with only three types (/b p pʰ/), i.e., losing the voiced aspirate /bʰ/ (still found in languages like Hindi). Commented Nov 21, 2024 at 1:24
17

Yes, there are – though the number of candidates differs quite a bit depending on the exact definition of your question.

Aspiration only

If you focus exclusively on aspiration, then a great number of Germanic languages, including English, have such a distinction.

In English (primarily European, as the question is about Europe, but also American and Antipodean), the contrast between /b/ and /p/ in initial position (i.e., at the very beginning of a stress unit, which is where the contrast is strongest) is primarily one of aspiration: /b/ is unaspirated, /p/ is aspirated.

Very often, /b/ is also voiced, but this is far less consistent than the aspiration of /p/. It’s not at all uncommon for a word like ban to be phonetically [paˑn] rather than [baˑn], but pronouncing pan as [paˑn] instead of [pʰaˑn] would almost certainly result in the listener hearing ban instead of pan, as very clearly demonstrated by Geoff Lindsey in Speech is really SBEECH on YouTube.

This tendency is even more pronounced in other Germanic languages like German, Icelandic and Danish where there’s less variation in whether or not /b/ is voiced.

  • The situation in German is somewhat complicated depending on dialect, but at least in many dialects the /b/ vs /p/ distinction is between primarily voiceless unaspirated [b̥] and primarily aspirated [pʰ].

  • Things are more straightforward in Icelandic, where the contrast is between [p] and [pʰ] (i.e., both are fortis and aspiration is the only distinguishing feature).

  • In Danish, both /b/ and /p/ are traditionally described as lenis and unvoiced, i.e., [b̥] and [b̥ʰ], though they’re often denoted [p] and [pʰ] for ease, since there is no lenis–fortis distinction.

  • Apart from these Germanic languages, the same type of aspiration-based distinction is also found in the extant Celtic languages, such as Scottish Gaelic (and to a lesser degree also in Irish Gaelic, though the unaspirated stops are often voiced in Irish – the situation there is similar to English) and Welsh (also similar to English). Breton is the only one where voicing is truly the main distinguishing feature, probably due to French influence.

Aspiration + voicing

All that is taking your question very literally: it’s the same distinction as Classical Greek π [p] vs φ [pʰ], leaving aside the rest of the phonemic inventory.

What it fails to take into account is that Greek additionally had a fully voiced counterpart β [b]. None of the Germanic or Gaelic languages have that. In fact, finding indigenous European languages that possess this feature is quite hard – I obviously don’t have intimate knowledge of all languages spoken in Europe, but I can only think of very few modern-day candidates that fit the bill, and they’re all located around the very edges of Europe (and apart from Armenian, they’re also all non-Indo-European languages).

  • Classical Armenian had a three-way distinction just like Classical Greek. Modern Western Armenian only has /b/ and /pʰ/ (a situation that rather bizarrely came about not through a simple merger of /p/ and /pʰ/, but by first merging /b/ and /pʰ/ and then subsequently voicing /p/ to /b/), but Eastern Armenian has retained all three.

  • Georgian almost fits: it has a three-way distinction which includes aspiration, but there is no plain /p/; instead, there is a voiced /b/, a voiceless aspirated /pʰ/ and an ejective /pʼ/, so aspiration isn’t the only distinguishing feature.

  • But some of the Sámi languages, like North Sámi, do seem to retain a fully ‘compliant’ /b ~ p ~ pʰ/ distinction (others, like Skolt Sámi only have a two-way distinction, usually voiced–unvoiced).

There may be more, but these are the only ones I can think of off the top of my head.

1
  • Comments have been moved to chat; please do not continue the discussion here. Before posting a comment below this one, please review the purposes of comments. Comments that do not request clarification or suggest improvements usually belong as an answer, on Linguistics Meta, or in Linguistics Chat. Comments continuing discussion may be removed. Commented Nov 29, 2024 at 8:26
4

Armenian (which is right next to Georgian) has a three-way distinction between voiced, voiceless and aspirated.

1
  • 1
    This varies by dialect. The system you describe is found in Classical Armenian, Agulis, South Eastern Armenian, and Yerevan (Middle Armenian and Karberd also have this system, but have t & d the other way round from the others). The dialects of Karin, Sebastia, Istanbul, Malatya, Van, Artsakh, and South Western Armenian do not have such a system (most have various two-way distinctions, but Karin & Sebastia have a three-way distinction between voiceless aspirated, voiced, and voiced aspirated) Commented Nov 21, 2024 at 11:08
-1

In English, "night rate" is pronounced differently from "nitrate."

1
  • That doesn’t really mean much, because those have different syllable structures: /naɪt.reɪt/ vs /naɪ.treɪt/; aspiration is always reduced (sometimes to the point of being completely lost) in coda position in English, but the sounds that differ between these two tokens are in fact the same phonemes. In order to show that aspiration is phonemic, you need to find minimal pairs with the same syllable structure. Commented Nov 23, 2024 at 18:05

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.