Periods after abbreviations

[Arnold M. Zwicky]

On Jun 29, 2004, at 5:57 AM, Jesse Sheidlower wrote:

> On Mon, Jun 28, 2004 at 11:18:17PM -0400, Baker, John wrote:
>>
>>         Jesse did not clarify why he thinks that "Ms" should
>> not have a period.  My initial thought was that he meant
>> that actual words, such as Ms and Mrs (neither of which is
>> an abbreviation for anything), should not have periods.
>
> "Mrs" isn't an abbreviation? Could have fooled me. I trusted
> my trusty OED when it told me that it's shortened from
> "mistress".

historically, yes.  but now it's a short spelling that is
conventionally (and, to some degree, arbitrarily) associated with the
pronunciation /mIsIz/, which is sometimes spelled out as <missus>.

a core, straightforward instance of an abbreviation is a short spelling
which can be converted to a fully spelled word by the addition of
letters and which gets its pronunciation from the pronunciation of the
longer spelling and its meaning from the meaning of the word with the
longer spelling.  <Prof.>, <Rte.>, and <fn.> are instances.

but the world of spelling is full of all sorts of other phenomena which
deviate in various degrees from the core instances.  <Mrs.>, pronounced
/mIsIz/, is one of them.  <fnn.>, pronounced /fUtnots/, is another.
<lb.>, pronounced /pawnd/,  is still another.  <MemChu>, pronounced
/mEm cu/, is yet another.  <IBM>, pronounced /ay bi Em/, is yet
another.  all the way up to <&>, pronounced /aend/.  there are many
different types of short spellings, and there is no natural way to
carve out categories other than the individual types (among them the
core type) or the whole set of short spellings.  you can, if you want,
reserve "abbreviation" as a technical term for the core type (in which
case, neither <Mrs.> nor <Ms.> is an abbreviation), or you can reserve
it as a technical term for short spellings as a whole (in which case,
both are abbreviations).  or you can give a disjunctive list of types
for which you will reserve it as a technical term (perhaps on the
grounds of historical derivation, though that would pick up <Miss> as
well as <Mrs.>, or on the grounds of usage in everyday language),
though then your list, as a definition of a *technical term*, is no
more defensible than any random list someone else might draw up.

the problem is that the question "is X an abbreviation?" doesn't make
sense unless you're clear about what counts as an abbreviation.  and
*that* question has two clear and defensible answers, a very narrow one
and a very broad one, plus any number of ad hoc answers; the answer you
give to the second question determines the answer to the first.

arnold (zwicky at csli.stanford.edu)