7

English coronal stop deletion, or TD-Deletion, is a variable process whereby word final /t/ and /d/ in clusters are deleted.

soft -> sof

A phonological rule for TD-Deletion could be given as:

{t,d} -> 0/ C__#

It can apply in a variety of morphological conditions.

Monomorphemes:

soft -> sof

Regular Verbal Morphology:

missed -> miss

Irregular Verbal Morphology:

kept -> kep

What I'm wondering is if there is any other good examples of a process like this in other languages. Here are my criteria.

  1. The process is easily describable, phonologically.
  2. The process applies variably.
  3. The process applies in a few different morphological contexts.
  4. Ideally, the process applies to a few different morphological exponents.

It would also be nice to know if the rate of variation is affected by the morphological structure it applies in/to.

JoFrhwld
  • 1,846
  • 15
  • 20
  • 1
    I don't understand under which conditions the TD-deletion is done. – Phira Sep 13 '11 at 21:24
  • @thei, does my edit clarify? – JoFrhwld Sep 13 '11 at 21:27
  • Maybe you could give one example? – Phira Sep 13 '11 at 21:29
  • 2
    There are 3 examples in the question: – JoFrhwld Sep 13 '11 at 21:30
  • I mean: When is the t of soft deleted? (And yes, I realize that it is deleted in clusters, I just don't know what constitutes a cluster.) I am not asking these questions to annoy you, I am quite interested and slightly ignorant on the subtler points of English pronunciation. – Phira Sep 13 '11 at 21:32
  • 1
    It's completely variable, so sometimes it deletes and sometimes it doesn't. Some factors, including the following word, can affect the probability of deletion, but there's nothing that absolutely determines when it will happen. – JoFrhwld Sep 13 '11 at 21:34
  • People are not hidden Markov models - they don't mentally roll a dice to determine which of multiple forms they'll produce. You simply don't have enough data to determine the right conditions for deletion. – Alek Storm Sep 14 '11 at 00:09
  • @AlekStorm: But it is possible that these conditions should fall outside the scope of phonology, isn't it? – Cerberus Sep 14 '11 at 00:59
  • 1
    @thei: Examples that work for me are /pæsbaɪ/ for "passed by" and /wɛsːaɪd/ for "west side". – tdhsmith Sep 14 '11 at 01:06
  • @Cerberus: I assume you're referring to sociological effects like discourse context and setting, which can cause speakers to modify their speech to fit a "higher" register. However, when controlled for these effects, t-deletion should still not vary probabilistically. – Alek Storm Sep 14 '11 at 01:14
  • 1
    @AlekStorm, there is such a thing as a variable process, and TD Deletion is one of them. To say there isn't is ignoring all of the sociolinguistic literature beginning with Weinreich, Labov & Herzog 1968. Variation has been found in all levels of grammar, from phonology, morphology, and syntax (kroch, 1989). – JoFrhwld Sep 14 '11 at 13:17
  • Is it important that it is in other languages than English? – Mitch Sep 14 '11 at 16:28
  • I think these separate questions should be asked separately. – hippietrail Sep 14 '11 at 17:35
  • @JoFrhwld, I'll concede that I am not a sociolinguist, and do not have immediate access to that work. Does it show specifically that within a speaker, when constrained to one social context, processes such as T-deletion vary probabilistically? – Alek Storm Sep 14 '11 at 17:37
  • 1
    @AlekStorm, within a speaker, within a social context (stylistic, etc.), within the same preceding and following segmental context, the deletion of /t/ and /d/ is describable only probabilistically. – JoFrhwld Sep 14 '11 at 18:53

2 Answers2

3

Would Spanish /s/ deletion fit? It applies in these morphological contexts:

  • Stems ending in /s/
  • The plural ending for nouns /-s/
  • The verbal conjugations for second person singular /-Vs/ and first person plural /-Vmos/
  • To the irregular 3rd sg. present tense verb form es (from ser). It also applies to the 2nd singular and 1st plural forms of this verb eres and somos respectively. (Whether these are represented as irregular root + usual affix or just an irregular stem is probably a matter of discussion.)
  • In some dialects, the 2nd sg. preterite marker /-Vste/ has a non-standard variant /-Vstes/ (by analogy with the other tenses). Insofar as this variant is produced, /s/ deletion can apply to it. (It would be tricksy to measure this, though, as after total /s/ deletion this variant is homophonous with the standard form. But you could count lenited tokens.)
Aaron
  • 3,612
  • 22
  • 25
1

I have three possible examples that might fit your criteria, two from English and one from French:

  • word final 's[unvoiced stop]' when followed by they noun plural morpheme or the verb singular morpheme 's':

-sps, -sts, -sks -> -ss ( long 's'?)

As in 'wasps' -> 'wass' or 'basks' -> 'bass'

More careful or slower pronunciation will pronounce all three phonemes, but usual conversation uses the rule.

  • '-ty-' -> '-tʃ-' (palatal glide to palatal affricate)

As in 'Don't you ...' -> 'Dontcha ...'. This seems to have no restriction in English to particular morphology (I've heard 'Rightcheer' for 'Right here').

  • In French, it is common to drop word final syllabic liquids:

'table' -> 'tab', 'autre' -> 'aut'

I can't think of any morphological interacting instances.

Mitch
  • 4,455
  • 24
  • 44