minimal pairs

Robert Whiting whiting at cc.helsinki.fi
Sun Oct 15 11:59:10 UTC 2000


On Wed, 04 Oct 2000 Larry Trask <larryt at cogs.susx.ac.uk> wrote:

>Max Wheeler writes:

>[on non-verbs in final edh]

>> Scythe? Lathe? Booth? Swathe? Tithe? Hythe (placename)? Blyth (placename)?
>> These old words perhaps undermine the hypothesis.

>Not all of these work for me, but most of them do, and I can add
>the noun 'edh' (or 'eth'), another name for barred-d.

There is also 'smooth' which forms an adjective/verb pair, both
with final [dh].

Not all of these work for me either, principally 'booth', which I
know only with final [th] (my dictionary says that final [dh] is
"esp[ecially] Brit[ish]").

But scythe is quite regular, the OE form being 'si:the' and hence
the <th> is intervocalic and should have gone to [dh].  My
pronunciation, however, is [sai] and hence is homophonous with
'sigh'.

'Lathe' has a counterpart with final [th] 'lath', but it is not
at all sure that the two are related.  Both are nouns and 'lathe'
is usually considered a loan.  Both also have unmarked correlate
verbs, 'lath' "to provide with laths," and 'lathe' "to turn on a
lathe."

'Swathe' (or 'swath') as a noun meaning "(like) a path cut with a
scythe" is difficult because it is mixed up with an unrelated
verb 'swathe' "to wrap, bind" (connected with 'swaddle').

'Tithe' is a frozen form, originally the same as the ordinal
number 'tenth' (OE 'te:otha' < 'teogotha').  Both verb and noun
are the same and only with [dh].

I won't comment on the place names, but I will point out that the
adjective 'blithe' has both pronunciations, with final [th] and
with final [dh].  The pronunciation with [dh] is regular since
the OE form was 'bli:the'.  The adjective 'lithe' is similar in
all respects.

>> I believe all /-Vth/ words are non-verbs (unless you include
>>"hath", "doth"), and nearly all are nouns (but for "with" in
>>some dialects).

An exception is 'quoth' "said" which is marginalized and has a
defective paradigm.  It is used only in the past tense, usually
with only the first or third person, and regularly has VS word
order ("Quoth the Raven, 'Nevermore'").  Unlike 'hath' or 'doth'
the <th> of 'quoth' is part of the root (< OE cwethan).  It is
not related to 'quote' which is a loan from Latin.  It is,
however, related to bequeath (or bequeathe), which can have
either [th] or [dh].  The corresponding noun, 'bequest', is also
irregular, having been influenced by 'quest' (a French loan).
So the pattern is truly broken on this one.

There are only a few /-Vth/ words that are not nouns:  'beneath',
'both', 'underneath', and, as pointed out, 'with'.  'Beneath' and
'underneath' are compounds, made up of 'be-' and 'under-'
combined with 'neothan' "below."  The latter element is related
to 'nether'.

>Agreed, except that certain nouns in final theta can also be
>used as verbs.

>A common example is the verb 'pith', as in 'pith a frog' (in a
>biology lab).  This is derived from the noun 'pith'.

>A second occurs, a little marginally, in the verb 'mouth off',
>which, in my experience, always has theta.

There is also 'bad-mouth', which, however, my dictionary lists
as either [th] or [dh].  In my experience, 'mouth off' can have
either [th] or [dh] as well.  My dictionary also records the verb
'mouth' as occurring with either [th] or [dh] although this is
recording general 'mouth' with [dh] and 'mouth off' with [th]
since it does not have a separate entry for 'mouth off'..

>A third is the peculiarly British use of 'bath' as a verb, as in
>'bath the baby'.  This, I believe, is unknown in American
>English, and perhaps in all other varieties of English.

>A fourth is the verb 'sheath', which is now rather common as an
>alternative to the traditional 'sheathe'.

>An extremely marginal fifth, which strictly only qualifies here
>in non-rhotic accents, is the British use of 'earth' as a verb,
>as in 'earth the TV set' (= US 'ground the TV set').

It is quite true that nouns in /-Vth/ can be used as verbs
without any change.  What we have here is the productive rule of
verb formation (NOUN --> VERB  / ...) coming into conflict with
the markedness rule for verb formation (final spirant voicing
rule).  As John McLaughlin has gone to some pains to point out,
"the" productive rule simply uses the noun as a verb without any
modification. The marked noun/verb pairs (with final spirant
voicing, which have all been in the language for a long time) are
beginning to contrast with unmarked noun/verb pairs created by
the productive rule.  I suspect that the phonemicity of [th] and
[dh] will manifest itself through the resolution of such
conflicts.  When the verb of the unmarked noun/verb pair and the
verb of the marked noun/verb pair have distinct meanings and one
always has only [th] and the other always has only [dh], then
this will be a phonemic distinction. So long as one verb or the
other can have either pronunciation, the distinction is not
phonemic.

This process, variation leading to change, is an extremely common
ingredient of linguistic change.  When two forms, created by
different rules, are in competition, a number of things may
happen:

1) Both forms will continue in use as free variants (thus 'dived',
   'dove' [verb]).

2) One form will simply replace the other and the other will
   disappear ('doth', 'does').

3) One form (normally the one formed by the productive rule) will
   become the general form and the other will be marginalized and
   used in a restricted sense ('brothers', 'brethren').

4) The two forms will split, each specializing in some meaning
   and become two separate lexical items.  The plural of 'staff'
   ('staf') was originally 'staves'; then a second plural
   'staffs' develops.  'Staves' then acquires a separate meaning
   from 'staffs' and a singular is formed by reversing the
   productive rule for plural formation:  'stave'.  The result is
   two lemmata with (slightly) different meanings, 'staff' and
   'stave'.  Only in the area of musical terminology are the two
   words coterminous.

>As an aside, I have just noticed that OED2 cites the very
>obscure noun 'mouthing', meaning 'entrance to a mine', with
>theta only -- perhaps somewhat unexpectedly.

I don't know what to do with this one either.

>Another aside.  While a blacksmith's shop is a 'smithy', with
>edh, a person called Smith gets the diminutive 'Smithy', with
>theta, in my experience -- just as in 'Kathy'.  (By the way, does
>anyone want to claim that 'Kathy' is less than fully English?)

My dictionary records 'smithy' with either [th] or [dh] and in my
dialect the hypocoristicon for a person named Smith is 'Smitty'.
But if you use [th] and [dh] to distinguish between a person
named Smith and a blacksmith's shop, then /th/ and /dh/ are
phonemes in your dialect (and this can't be argued against; de
gustibus non est disputandum).  But it isn't in mine.

And 'Kathy' is fully English in the same way that 'sympathy' and
'pathetic' are.  Although 'C/Katherine' (and probably the
diminutive 'Kathy') has been in the language longer, they all
postdate the sound change that voiced intervocalic spirants.  So
'Kathy' follows the same rules, as English as it may be.  The
question is does [kathi:] contrast with [kadhi:] anywhere?

The point is not that English words can't have intervocalic [th].
The point is that all Greek and Latin loans and neologisms in
English have intervocalic [th].  There are a number of examples
of intervocalic [th] in native English words caused by such
things as the position of the stress which caused the voicing
rule not to operate or the fact that there was originally a
doubled [th] in the word which was subsequently simplified
leaving an intervocalic single [th] high and dry.  But the point
is that the distribution of intervocalic [th] in English is not
arbitrary.  It can always be accounted for by rule.  This is why
English can get by with a single graphic symbol <th> for both
[th] and [dh].  Even when the English alphabet had separate signs
for both sounds, þ (thorn) and ð (edh), they were never used
systematically, but were more or less interchangeable.  The
pronunciation was always predictable.

Bob Whiting
whiting at cc.helsinki.fi



More information about the Indo-european mailing list