LINGUIST List 6.1672

Wed Nov 29 1995

Misc: Unusual Sound Change t > h, Gender in Linguistics

Editor for this issue: Ljuba Veselinova <>


  1. benji wald , unusual sound change t > h ?
  2. "Crookston, Ian [HSC]", submission to linguist

Message 1: unusual sound change t > h ?

Date: Sat, 25 Nov 1995 23:18:00 unusual sound change t > h ?
From: benji wald <IBENAWJMVS.OAC.UCLA.EDU>
Subject: unusual sound change t > h ?

Unusual sound change. I don't know. But is
 t > h
an example?

I only know of one area where this has happened. Well, two, but not too
far apart.

1) Most of the Sabaki group of the Bantu family in coastal Kenya.
cf. Swahili watoto watatu (children three)
 Giriama ahoho ahahu 'three children'

2) Separately among the Chagga dialects of the Kilimanjaro region of
Tanzania, also Bantu. Siha, one of the most western dialects of
Chagga has t > h.

>From various types of evidence, it is pretty clear what the intermediate
stages in the t > h sound change were. Crucial is the intermediate
change from /t/ to a *voiceless* /r/-like flap. In the rural dialects
of Swahili, e.g., Vumba, and also the off-shore Comorian languages,
this /r/ voices and merges with /r/ from other sources, e.g.,
 Vumba waroro wararu

In fact, the self-stereotype of Vumba refers to this peculiarity.
 rere rwaa (bring take -- imperative forms)
which contrasts with urban lete twaa.

 Note then that it is not only morpheme-internal intervocalic, but
also word-initial -- generally also intervocalic because any preceding
word would end in a vowel in the relevant Bantu area. The only hint
that the change may have originated in morpheme-internal intervocalic
position is that the voiceless flap is an allophone of /t/ in the
urban Swahili of Mombasa, where it varies with /t/, e.g., /pata/ or
/paRa/ 'get' cf. Vumba vara. That allophone never occurs for morpheme
initial /t/.

In Giriama and most of the rest of the Sabaki group, it seems that
point of articulation was simply lost, so that the voicelessness of
the flap alone remained, which is /h/.

The change seems to have taken place independently in Chagga,
separated from the coast by an extensive conservative /t/ area. But
then in Chagga all kinds of things happen to /t/ according to
dialects. The Central dialects have an /r/-like sound, but often with
spontaneous palatalization, so the outcome of /t/ to flap is identical
to the Czech r with the hachek. This is characteristic of the
well-known Vunjo and Moshi dialects. Further West, the Machame dialect
velarises instead, so /t/ ends up sounding like a voiceless Parisian
French /r/, very /x/-like. Finally further west in Siha this is all
smoothed to /h/. The areal continuum, then, makes it look like the
route in Siha is different from the one taken by Sabaki, because it
may be that it passed through the velar stage still preserved in
Machame. There is no evidence for such a stage in Sabaki.

It did not strike me at first that this might be an "unusual" change
since I am so familiar with it, but then I thought about it and don't
know any other languages anywhere where such a change has taken place.
Also when I've asked linguists "how do you suppose the change t > h
takes place?" they have never guessed a flap intermediary, but tried
to imagine an aspirated t becoming h, something which definitely does
not happen in the relevant area (aspirated t remains aspirated t), and
is not based on known examples to the linguists I have asked.

So, of course, I would be interested if there are recurrences of t > h
elsewhere in the world -- by any route -- and find it interesting that
the change occurred independently (I have no doubt) in two areas that
are relatively near each other.

To finish off discussion here, the coincidence in the two areas of t >
h is partially explained by general lenition processes in the larger
area which connects the Chagga and Sabaki. The change also affects
/p/, cf. Vumba vara (where /v/ is a bilabial fricative varying in
voicing) from pata. In Southern Sabaki a voiced bilabial is the usual
outcome, but in Northern Sabaki /h/ is the outcome. cf. Digo Southern
uvevo Giriama Northern uheho (Swahili upepo) "wind". In both kinds of
Sabaki t > h is the invariable outcome. t > h also affects Southern
Pokomo (north of Giriama) but only marginally p > a voiceless bilabial
affricate. Thus in this area, t > h is the more general and areally
extensive change.
 -- Benji

> 3)
> Date: Sat, 18 Nov 1995 14:50:09 EST
> From: ("R. Joe Campbell")
> Subject: Another Unusual Sound Change
> In the Nahuatl spoken in some of the towns of Guerrero, Mexico (e.g.
> San Augusti'n Oapan and Xalitla), the double 'l' sequence
> (morphophonemically and phonetically double) /ll/ in other dialects is
> realized as [hl]. The segment [h] that occurs word-internally before
> consonants in other dialects is deleted. It should be noted that S.A.O.
> shows no tendency to simply devoice /l/ (i.e., > voiceless lateral
> fricative) in syllable-final position generally (as is found "in some
> other dialects").
> S.A.O. "standard" [!?] gloss
> xahli xalli sand
> otli ohtli road
> 218 Ridgeview Drive
> Bloomington, Indiana 47401
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: submission to linguist

Date: Thu, 23 Nov 1995 12:09:00 submission to linguist
From: "Crookston, Ian [HSC]" <>
Subject: submission to linguist

I note with interest Dick Hudson's summary on gender in linguistics. I
experience the extreme case: speech & language therapy training which
in the UK is about 95% female - far higher than any paramedical

I think I can offer some tentative explanations as follows.

Why do so few men apply for S&LT training? Probably because the pay
ceiling for the profession is known and is low. In general
linguistics, there might be a related factor: job prospects for
linguistics graduates might be perceived as comparatively uncertain.

Why do so many women apply for S&LT training? (It _is_ extremely
popular). I suggest it is because they don't know what it
involves. When they arrive and find out there is a certain feeling of
shock. I'm suggesting that these young women, having been turned off
science perhaps by the absence of female science teachers, are looking
to a linguistic discipline as an escape from science. Happily, they
tend to get over the shock and approach the training with vigour and
success. In general linguistics too, surely there is a certain
tendency to take up the course without having a clear idea what it is.

Having said that, it must also be admitted that gender specialisation
is self-perpetuating. We had one good male student leave recently
simply because he couldn't face being part of such a tiny minority all
his life: I imagine that prospect puts off a certain number of
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue