Tom Pullman: Fonts with buailte characters

[Elizabeth J. Pyatt]

Date: Tue, 22 Apr 2003 15:23:55 +0100 (BST)
From: Tom Pullman <tjop2 at cam.ac.uk>
To: The Celtic Linguistics List <CELTLING at LISTSERV.LINGUISTLIST.ORG>
Subject: Re: Kevin Riley: Fonts with buailte characters

On Tue, 22 Apr 2003, Kevin Riley wrote:
>
>>  >>     It should be pointed out that the system with 'h's also
>>  >>     goes back to the 17th century, perhaps even earlier...
>>  >>
>>  >>           le meas,  Doireann
>>  >>
>>  If my memory is correct, the use of 'h's is the original system, with the
>>  superscript dot [punctum delens] being restricted to the 'f' as it became
>>  silent.  The use of 'h' was borrowed from similar conventions used in
>>  Romance and British languages.  Nasalisation was not originally marked at
>>  all, but was left to the reader to supply.
>
The original Old Irish system used <h> to mark lenition, but only on
consonants which the scribes were used to seeing followed by <h> in Latin
(mostly in Greek loanwords) - i.e. <c>, <p> and <t>. The "punctum delens"
was used for <s> as well as <f> in some circumstances, notably after
leniting forms of the article. In other words, both <h>-insertion and the
superscript-dot existed in the oldest form of the orthography. Lenition
was never marked on <b>, <d>, <g>, <m>, <n>, <l> or <r>, all of which
consonants could be lenited in Old Irish.

Obviously it was desirable to mark lenition on more consonants than the
initial system allowed for. Both of the above strategies have been
extended (in different orthographic systems) to cover <b>, <d>, <g> and
<m>, and also the consonants originally covered by the competing method.

It is notable that (word-inital) lenition has never, as far as I know,
been marked orthographically on <n>, <l> or <r>, despite the fact that the
difference in pronunciation between the tense and lax (= non-lenited and
lenited, when word-initial) varieties of the consonants was large enough
to merit different spellings word-internally (<nn>/<n>, <ll>/<l>,
<rr>/<r>). Possibly it was felt that both the superscript-dot and
<h>-insertion ought to result in a graph representing a fricative or zero,
as this is the case for all of <bh>, <dh>, <fh>, <gh>, <mh>, <ph>, <sh>
and <th> - or at least it was before various sound changes. /p/ > /f/
(etc.) is an intuitively different-feeling change to /L/ > /l/. It is
possible, therefore, that neither of the existing strategies was thought
acceptable for this reason.

That left only the double-graph/single-graph alternation for writers to
play with, and that was clearly unacceptable for use word-initially: not
only did it clash radically with familiar Latin orthographical conventions
and make the base form longer than the lenited form*, but in the context
of a spelling reform for clearer indication of lenition it would have
necessitated a re-spelling of the initial consonant of the base form of a
huge number of words. Scribes used to writing <lia>, <lón>, <ní>, <na>,
<rún> etc. would have to alter them to the unfamiliar and odd-looking
<llia>, <llón>, <nní>, <nna>, <rrún>. The undesirability of this would
have been compounded by the fact that doubling of initial consonants,
these ones especially, was used at some stages of the language to
represent nasalisation. (Another story - but nasalising prefixes and words
may have caused gemination of liquids and nasals in early forms of Irish,
which I assume is the justification of the spelling. In fact, the history
of the representation of nasalisation appears to show a similar pattern
of competing strategies to that of the representation of lenition.)

Thus in Modern Irish and Scottish Gaelic we have a system which still does
not mark initial mutation of /L/ and /N/, after a thousand years and more
of evolution of the spelling system.

Tom

*Not necessarily a problem, of course - c.f. Welsh <ll> and <l>.

--

Tom Pullman			¦
Gonville and Caius College	¦	 Whenever I speak Tlingit
Cambridge CB2 1TA		¦	I can still taste the soap
tjop2 at cam.ac.uk			¦

--
=-=-=-=-=-=-=-=-=-=-=-=-=
Elizabeth J. Pyatt, Ph.D.
Instructional Designer
Education Technology Services, TLT
Penn State University
ejp10 at psu.edu, (814) 865-0805

228A Computer Building
University Park, PA 16801
http://www.personal.psu.edu/ejp10
http://tlt.psu.edu