Page 1 of 1

Affricates

Posted: Fri Jan 12, 2018 2:55 pm
by yangfiretiger121
Is there a specific reason articulation of combinations like [pɹ] as an affricate is judged to be impossible while /d͡ɹ̝/ is considered an affricate?

Re: Affricates

Posted: Fri Jan 12, 2018 3:03 pm
by WeepingElf
yangfiretiger121 wrote:Is there a specific reason articulation of combinations like [pɹ] as an affricate is judged to be impossible while /d͡ɹ̝/ is considered an affricate?
In an affricate, both components are homorganic, i.e. produced by the same articulator. [pɹ] combines a labial with an alveolar sound, thus no affricate. (Also, [ɹ] is not a fricative.)

Re: Affricates

Posted: Fri Jan 12, 2018 10:00 pm
by Nortaneous
can have unit rhotacized peripheral plosives (Pumi iirc), they're just not affricates

Re: Affricates

Posted: Sun Jan 14, 2018 4:12 pm
by vergil
How does something like [tʙ] feature into this homorganicism?

Re: Affricates

Posted: Sun Jan 14, 2018 7:54 pm
by Travis B.
vergil wrote:How does something like [tʙ] feature into this homorganicism?
That is not homorganic.

Re: Affricates

Posted: Mon Jan 15, 2018 3:43 am
by Nortaneous
Travis B. wrote:
vergil wrote:How does something like [tʙ] feature into this homorganicism?
That is not homorganic.
right, but it's still sometimes called an affricate.

Re: Affricates

Posted: Mon Jan 15, 2018 1:29 pm
by Travis B.
Nortaneous wrote:
Travis B. wrote:
vergil wrote:How does something like [tʙ] feature into this homorganicism?
That is not homorganic.
right, but it's still sometimes called an affricate.
Probably the best measure is whether something can slot into the syllable structure as a unitary entity, or whether something has to be treated as a cluster with regard to syllable structure and syllabification, metricality, and so on. Unfortunately, making this determination is not always possible (e.g. languages with complex onsets and codas in the first place where syllable weight has little consequence so thus cannot be used as a distinguishing measure).

Re: Affricates

Posted: Mon Jan 15, 2018 4:59 pm
by Sumelic
vergil wrote:How does something like [tʙ] feature into this homorganicism?
[tʙ] seems to be very rare. I don't know much about the languages where it occurs, but "Linguistics: An Introduction (William B. McGregor, 2015) describes it as an "unusual coarticulated phone".

The Wikipedia article "bilabial trill" says:
There is [...] a very rare voiceless alveolar bilabially trilled affricate, [t̪͡ʙ̥] (written ⟨tᵖ̃⟩ in Everett & Kern) reported from Pirahã and from a few words in the Chapacuran languages Wari’ and Oro Win. The sound also appears as an allophone of the labialized voiceless alveolar stop /tʷ/ of Abkhaz and Ubykh, but in those languages it is more often realised by a doubly articulated stop [t͡p]. In the Chapacuran languages, [tʙ̥] is reported almost exclusively before rounded vowels such as [o] and [y].
I also found an article called "Prestopped bilabial trills in Sangtam", by Alexander Coupe, that refers to /t͡ʙ/ and /t͡ʙ̥ʰ/ using the phrasing in the title, and puts them in a separate row from the affricates in the table of Northern Sangtam phonemes.

It does seem that "tʙ" has been referred to as an affricate by at least some people. I'm not sure if that application of terminology is a good idea. For one thing, I don't know if it's appropriate to think of the "ʙ" element as a fricative (I don't know what phonetic criteria can be used to rigorously differentiate between trills and fricatives; they seem to be somewhat similar phonetically, as seen in the historical or synchronic relationships between the trill [ʁ] and the fricative [ʀ], or the trill [r] and a fricative like [ʐ] or [z], in a number of languages), although I don't know if I might be committing the etymological fallacy here by supposing that the definition of "affricate" should be related to the definition of "fricative". I have never seen coarticulated stops like [k͡p] or pre-stopped nasals like [ᵈn] called affricates, though, so it seems like the term must be in some way more specific than just "something that is phonetically like a cluster but phonologically acts like a single phoneme".

It seems to me like one fairly convincing argument for calling /t͡ʙ/ an affricate in some particular language might be if the language has normal affricates like /ts/, /dz/ or /tʃ/ that are phonologically distinct in some way from other obstruents, and /t͡ʙ/ phonologically behaves more similarly to the to the affricates than to any of the other obstruents, but I haven't found any examples so far in my brief bit of Googling. It sounds like it may actually pattern like a labialized plosive in some of the languages in which it occurs. That said, according to Coupe, the development of bilabial trills may be associated with high rounded vowels in particular, which does seem to suggest a possible link with affricates in that high vowels may also be a conditioning factor for affrication like [t] > [t͡s]. Still, it seems to me that the similarity between conditioning environments for affricate development and trill development wouldn't necessarily require that we use the same terminology for the resulting sounds. From what I remember, high vowels can be a conditioning environment for aspirated consonanants for similar reasons, but we don't call normal aspirated consonants "affricates".

yangfiretiger121, I'm confused by your statement in the original that "combinations like [pɹ] as an affricate is judged to be impossible". If you don't know of any reason for this to be true, why do you think it is true? Did you read it somewhere? If so, it would be nice to know where. Academic sources often, although not always, have explanations or citations to support statements like this.

The Wikipedia article on the the Osage language actually presents an interesting possible counterexample: it lists /br/ in a single cell of the consonant table, although a later note refers to it as a cluster ("The /br/ cluster also depends on dialect. It is sometimes pronounced [bəl] or [bər]"). It doesn't seem to contrast with /pr/ though, and a note mentions that Osage lenis plosives may be "lightly voiced" in at least some circumstances, so it could probably be analyzed synchronically as a cluster /pr/ where the /p/ is realized as due to regressive voicing assimilation (diachronically, it apparently mainly comes from w + r, which suggests an alternative synchronic analysis as /wr/).

Re: Affricates

Posted: Wed Jan 31, 2018 1:33 pm
by yangfiretiger121
I was looking at the greyed-out cells in the sibilant affricate/fricative rows of Wikipedia's IPA chart, in which the color means "articulation has been judged impossible." Then, I realized that [pɹ], etc. would be non-sibilant because of the [ɹ]. On top of that, those positions are taken by [pɸ bβ], thereby voiding the question.

Re: Affricates

Posted: Mon Feb 12, 2018 4:03 pm
by vec
Coarticulate = heterorganic cluster that behave as single phoneme
Affricate = coarticulate of homorganic plosive and fricative

So, an affricate is a subset of coarticulates. Coarticulates must be 1) homorganic, 2) stop+fricative, in order to be defined as affricates.

Re: Affricates

Posted: Mon Feb 12, 2018 4:18 pm
by Travis B.
vec wrote:Coarticulate = heterorganic cluster that behave as single phoneme
Affricate = coarticulate of homorganic plosive and fricative

So, an affricate is a subset of coarticulates. Coarticulates must be 1) homorganic, 2) stop+fricative, in order to be defined as affricates.
One should distinguish phones where multiple POA are articulated simultaneously with affricates, where there is a plosive at one POA immediately followed by a fricative release at the same POA or a very close POA (e.g. German /pf/) which behaves as single unit for the purposes of phonotactics.