LINGUIST List 13.1566

Fri May 31 2002

Sum: Apostrophe Rules in Langs

Editor for this issue: Marie Klopfenstein <>


  1. Daniel Buncic, Apostrophe rules

Message 1: Apostrophe rules

Date: Fri, 31 May 2002 15:41:17 +0200
From: Daniel Buncic <>
Subject: Apostrophe rules

On 5 April (sorry for the long delay!) I posted a question
on the use of the apostrophe in various languages

I received 30 answers from Miikka-Markus Alhonen (on Finnish and Swahili),
Michael A. Covington (English, German), Peter T. Daniels (English, German),
Yehuda N. Falk (Hebrew), Christine Haunz (on an experiment with English
native speakers), Johannes Heinecke (Breton, Italian), Wolf Peter Klein
(German), Mark A. Mandel (transcriptions and transliterations, Hawai'ian,
Tahitian, English), Barbara Zurer Pearson (English), Marc Picard (English,
German), Lukas Pietsch (Turkish), Donald F. Reindl (Slovene), Charley Rowe
(English, German), Nino Vessella (Swahili, Italian), R�my Viredaz (English,
French, German), and Jeremy Whistle (French, Italian, Turkish).

My problem was the following: Usually it is assumed that the basic function
of the apostrophe (in English, French, German, Italian, and a lot of other
languages) is the marking of elision. In many languages, however, the
apostrophe also marks boundaries where there is no omission; these cases are
usually treated as 'exceptions'. My thesis is that in fact most cases that
are commonly explained by omission can just as well be explained by the
'separative' function of marking (mostly morphological) boundaries, so that
instead of two functions of the apostrophe we get one. There are some few
instances left in which apostrophes occur in places where there is no
boundary, but they are confined to 'transcriptions' of colloquial speech or
poetically 'deformed' language and are not found in formal texts (see 3.

Consequently I have sorted the answers I received (with my personal
additions concerning German, English, Ancient Greek, Russian, Ukrainian,
Dutch, IPA, and shorthand) by functions of the apostrophe:
 1. boundary marker (exclusively)
 2. marker of boundary and elision
 3. elision marker (exclusively)
In addition to the two functions discussed above, there are three more
 4. abbreviation mark
 5. diacritic
 6. quasi-letter
In the following I will use the SAMPA conventions for phonemic

 The first example are ENGLISH possessive forms, which are used
predominantly with personal names, e.g. <Mary's car>, <John's kitchen>. In
possessive forms of nouns ending in an /s/, the apostrophe marks a zero
morpheme, e.g. <the Smiths' car>, <Athens' townhall>. Because of the
similarity of such forms to those mentioned before, there are many
misspellings like <Athen's townhall> (see below). -- Attempts to explain
this use of the apostrophe as elision are abstruse: Either there is claimed
to have been an <e>, which is "still" pronounced (but not written!) in
<Octopus's gardens> /-sIz/, or the possessive form is said to be derived
from dialectal forms with a possessive pronoun, viz. <John his kitchen>.
Nothing at all is missing in the possessive apostrophe of words ending in
/s/ like <the Smiths' car>. There has never been a sound after the /s/.
However, there is a zero morpheme: <Smith+s+0> 'Smith + plural +
possessive'. In the pronunciation of <Octopus's gardens> mentioned above the
apostrophe marks a place where there is not only nothing omitted but even
something added, namely /I/. -- Note that possessives are marked with an
apostrophe only if there is a morpheme boundary. Possessive pronouns, which
do not contain a boundary, do not have an apostrophe (*<hi's>, *<it's>
 ENGLISH plurals of "non-words or unlikely nouns or non-canonical words"
(Charley Rowe) like <ABC's>, <P's and Q's>, <5's and 10's>, <why's and
wherefore's> often have an apostrophe. Peter T. Daniels comments on such
forms: "These are not normal according to contemporary stylebooks: the
Chicago Manual of Style (the de facto American standard) uses 60s and _l_s
(i.e., the l is italic, the s isn't). An apostrophe is inserted only in
desperate cases."
 In DUTCH, plural forms of words ending in a vowel are written with an
apostrophe, e.g. <auto's> 'cars'.
 Historically, in ENGLISH, too, plural forms - especially of foreign
words - were written with an apostrophe, e.g. in the 18th century <idea's>,
<toga's>, <folio's>, <quarto's>,<genius's>, <species's> (quoted from
Elizabeth S. Sklar, "The possessive apostrophe: The development and decline
of a crooked mark", College English 38.2, 1976, 175-183, 178).
 Furthermore, in contemporary ENGLISH there is a widespread wrong use of
such forms as <thank's> (58,500 Google hits for this search term) or
<apple's>, where there is a morpheme boundary before the <s>, and <Athen's>
or <it's> (for <its>), where there is obviously felt to be a morpheme
boundary, in analogy to the possessive marker mentioned above.
 In GERMAN, the use of <'s> for genitive forms of names is usually viewed
as incorrect, but the orthography reform of 1998 sanctioned the common usage
(as a "can" rule, not a "must"!) in such instances as <Andrea's Boutique>
'Andrea's boutique'; the rules state correctly that <Andreas> as the
genitive form of <Andrea> (female name) could be confused with the
nominative form of <Andreas> (male name), but it is not clear whether the
apostrophe is allowed only in these cases of possible confusion or in
general. In practice it is used with any <s> ending and frowned upon by
those who consider themselves educated. Of course the popularity of such
forms is influenced by English, which gives them a certain 'cool value'.
(Charley Rowe claims that "you certainly seem to find it more in Berlin,
which is the apparent leader in English borrowing".) However, in fact the
genitive apostrophe is about as old in German as it is in English (attested
since the 17th century). Wolf Peter Klein presumes that historically the
genitive apostrophe developed from an apostrophe marking elision in
instances like <Gott's Wahrheit> for <Gottes Wahrheit> 'God's truth', and
that its function was re-interpreted as separating the stem from the ending,
which gave rise to the plural apostrophe too, as in <alle Comma's> (nowadays
only <alle Kommas>) 'all the commas' (quoted from one of Schlegel's letters
from 1804 in Gerhard Zimmermann, "Der Genitivapostroph im Deutschen",
Muttersprache 44, 1983/84, 417-434, 424). -- The 1998 orthography reform
also accepted the apostrophe in adjective forms like <das Ohm'sche Gesetz>
(alongside <das ohmsche Gesetz>) 'Ohm's law', where the suffix <sch> is felt
to be related to <isch>, but it cannot be replaced, *<ohmisch> does not
exist (and would have to have an umlaut if it existed: *<�hmisch>).
 A morpheme boundary is also marked in the wrong usage of GERMAN
imperatives which never had an /-e/ ending, e.g. <gib'> instead of <gib>
'give!', where *<gibe> has never existed in German. Of course, there is an
analogy to forms where there _can_ be an /-e/ ending like <gehe> alongside
with <geh> 'go!' (the spelling of which as <geh'> was correct until 1929 -
the 10th edition of the Duden orthographic dictionary). In this latter case
<geh> there is indeed elision, but not in the former case <gib>. In both
cases, however, there is a zero ending! -- Other instances of 'wrong' German
apostrophes (called "Dottikon" after Sven B�ttcher's translation of Douglas
Adams' and John Lloyd's _The deeper meaning of Liff_, _Der tiefere Sinn des
Labenz_, M�nchen 1992, or "Kapostroph", as in Philip Oelwein's
"Kapostropheum", are <H�us'chen> for
<H�uschen> 'small house' (which can be read as /"hOYSn/ instead of
/"hOYsCn/) and <das fit'e Sportstudio> for <das fitte Sportstudio> 'the
fit sports studio' (where the foreign word <fit> is hard to recognize
because of its <tt>). Both examples are drawn from Daniel Fuchs, "Die
Apostroph-S-Hass-Seite" ('The apostrophe-s hate page'),
<>, and both mark morpheme
 R�my Viredaz gives a FRENCH (non-standard?) example of an exclusively
separative apostrophe, the Senegalese surname <N'Diaye> or <n'Diaye>
(besides <Ndiaye>) "to avoid the un-French initial group <nd>".
 In TURKISH, flexives are separated by an apostrophe from the stem of
proper names (both foreign and native, e.g. <Ankara'dan> 'from Ankara'),
while derivational suffixes are separated only in foreign proper names (e.g.
<Lille'li> 'inhabitant of Lille', but <Bursali> 'inhabitant of Bursa').
Problems arise with foreign proper names that are not felt to be foreign
(i.e. that function as exonyms) like <Paris>: <Paris'li> or <Parisli>
'inhabitant of Paris'? (Thanks a lot to Lukas Pietsch, who gave me a German
summary of Nemiye Alpay's article "Dil Meseleri" in _Radikal_ of 11 April
 Similarly, in FINNISH the apostrophe is used when a foreign proper name
"which ends in a written consonant but a pronounced vowel" (Miikka-Markus
Alhonen) is inflected, e.g. <Bordeaux'ssa> /bordo:s:A/ 'in Bordeaux',
<Versailles'ssa> /versAis:A/ 'in Versailles' (but if <Versailles> is
pronounced /versAj/, i.e. ending in a consonant /j/, oddly enough, the
inessive form has to be <Versaillesissa> /versAjis:A/).
 In RUSSIAN, foreign words are usually transcribed, but if they are left
in their original Latin orthography, difficulties arise when Cyrillic
morphemes (inflectional, derivational or compositional) have to be added to
these words. The rule seems to be that derivatives (<DOS-ovskij> 'of DOS
(adj.)') and compounds (<IBM-sovmestimost'> 'IBM compatibility') are formed
with a hyphen, while inflection is added with an apostrophe: <laptop'ov> 'of
laptops (gen. pl.)'. [The second part of all these examples is to be
understood as originally written in Cyrillic letters, which had to be
transcribed here; for a Cyrillic illustration see]
 In POLISH, inflected foreign words are written with an apostrophe if the
inflection changes the pronunciation of the stem, e.g. <Harry'ego> 'of Harry
(gen. sg.)', pronounced without the <y> as /xarego/; but <Harrym> 'Harry
(instr./loc. sg.)' pronounced /xarim/ with the <y>.
 In HEBREW, sporadically suffixes borrowed from Yiddish (e.g. the
diminutive /le/) are preceded by an apostrophe, but this does not accord to
the standard orthography.
 In UKRAINIAN, the apostrophe marks a morpheme boundary before /j/, since
the orthography otherwise does not differentiate between the combination
[palatalized consonant + vowel] (appearing anywhere) and the combination
[consonant + j + vowel] (appearing only at the morpheme boundary), e.g.
<z'jasovuvaty> 'to explain' (pronounced /zja.../, but cf. <zjabra> 'gills'
pronounced /z'a.../), <ob'jednanyj> 'united', <vid'jizd> 'departure'
 In RUSSIAN, in the 1920s the apostrophe was used in the same way as in
Ukrainian, e.g. <ot'ezd> 'departure'. Nowadays, the so-called 'hard sign' is
used in this function (again).

 In SEVERAL LANGUAGES, years can be named without the first two figures,
and this can be indicated by an apostrophe, e.g. <'93> for <1993>.
 In ENGLISH <they've> for <they have>, <it's> for <it is>, <o'clock> for
"of the clock" etc. the apostrophe marks a morphological word boundary where
two words have been phonologically melted into one at the expense of one or
more phonemes, so that at the same time the apostrophe marks their elision.
Rarely the apostrophe also marks syllable boundaries: <Hallowe'en> for
"(All) Hallow(s') E(v)en", where only one of the three elisions is indicated
by an apostrophe and where the form <Halloween> without the apostrophe is
nowadays much more common, just as the pronunciation /-wi:n/ instead of
/-wi:n/; a syllable boundary is also indicated in <ma'am> for <madam>,
where at least historically there is also morpheme boundary: <ma dame> 'my
 In ITALIAN the apostrophe marks 'elision' at a boundary where the
uncontracted forms are ungrammatical: <l'amica> 'the friend (female)'
instead of *<la amica>, <l'albero> 'the tree' instead of *<lo albero> or
*<il albero>, <d'ogni> 'of all' instead of *<di ogni>, etc.
 The same happens in FRENCH: <l'ami> 'the friend (male)' instead of *<le
ami>, <c'est> 'this is' instead of *<ce est>. In some words the apostrophe
is now usually replaced by a hyphen, e.g. <grand-mere> for older
 In SWAHILI <n'na> for <nina> 'I have', <m'menunua> for <mumenunua> 'you
(pl) have bought', where <ni> and <mu> are prefixes in the function of our
personal pronouns.
 For GERMAN apostrophe use, the standard example in the Duden is <Das
Wasser rauscht', das Wasser schwoll> 'The water rushed, the water swelled',
which shows that the sense would change if you did not put the apostrophe
here (<rauscht> is present tense, while <rauscht'> for <rauschte> is past
tense). So basically what is indicated here is a zero morpheme. The
imperative forms like <geh'> instead of <geh> 'go!', where there is also an
alternative form <gehe>, have officially been outdated since 1929, but are
still very popular (cf. 1. above for further German imperatives).
 In FINNISH, the apostrophe indicates syllable boundaries when similar
vowels become adjacent due to the elision of [k], e.g. <vaaka> 'scale' + <n>
'genitive' becomes <vaa'an> 'of a scale'. In old texts (e.g. the "Kalevala")
this works with other consonants as well, e.g. <kahta'alta> 'from two
places', where the original dialectal form was <kahtahalta>, but the modern
standard form is <kahtaalta>. In "poetic language and in songs", the elision
of word final vowels is also indicated by an apostrophe (e.g. <Kaikk'
samett' silkkipuku> for *<Kaikki sametti silkkipuku>). This occurs not only
at the word boundary but also at morpheme boundaries inside compounds, e.g.
<joka> 'every' + <ainoa> 'only' = <jok'ainoa> 'every single'.
 In ANCIENT GREEK, the apostrophe marked the elision of vowels at the
word boundary, by way of which one of the words became a clitic, e.g.
<ap'emou> for *<apo emou> 'from me', <all'hyp'autou> for *<alla hypo autou>
'but by him'.
 In Early Modern ENGLISH, the apostrophe was used under the same
conditions as in Ancient Greek: "Christians d'obey th'officers and rulers,
that b'appointed of God in th'Earth" or "writ th'articles plaine
t'understand" (John Hart, "The opening of the unreasonable writing of our
Inglish toung", 1551, new edition 1951, 153, quoted in: Vivian Salmon,
"Orthography and punctuation", in: The Cambridge history of the English
language, ed. Roger Lass, Cambridge 1999, vol. 3, 13-55, 22 f.).

 ENGLISH contractions of /not/ have the apostrophe inside this morpheme:
<can't> for <cannot> (according to Peter T. Daniels written <ca'n't> until
the 19th century, e.g. in the 1865 edition of Lewis Carroll's _Alice in
Wonderland_), <isn't> for <is not>, <aren't> for <are not>; furthermore
there are some rare singular cases, such as <fo'c's'le> for <forecastle>.
Note that - in contrast to the possessive apostrophe - all these forms are
_not_ used in formal texts but only for 'recording' colloquial speech.
 A few GERMAN colloquial contractions have non-morphological apostrophes,
too, e.g. <'n Abend!> for <Guten Abend!> 'Good evening!'; this phrase cannot
even be capitalized at the beginning of a sentence, so that the spelling has
to be reformed anyway. Gallmann (_Graphische Elemente der geschriebenen
Sprache_, T�bingen 1985, 267) proposes <'N Abend> - but why not put the
apostrophe between the morphemes, which are pronounced as one word anyway -
 ITALIAN non-standard post-vocalic or syllabic apocope: <da'> for <dai>
'give!', <fa'> for <fai> 'do!', <sta'> for <stai> 'stay!', <va'>, <un po'>
for <un poco> 'a little'.
 In SLOVENE the sound /i/ is often dropped in colloquial conversation,
and this can be indicated in writing to convey the colloquiality; in many
cases the apostrophe thus stands at the morpheme boundary: <druz'ga> for
<drugega> 'of the other', <pr' Mladinsk' knig'> for <pri Mladinski knjigi>
'at the Mladinska Knjiga publishing house', <pr' an' his^'> for <pri eni
his^i> 'at one house'. However, in other cases the apostrophe is placed not
at the morpheme boundary, e.g. <'mam> for <imam> 'I have', <b'la, b'lo,
b'li> for <bila, bilo, bili> 'been', <tol'k> for <tolik> 'such'. All these
forms are sometimes written without the apostrophe as well, and neither
representation is sanctioned by official orthography.
 In FRENCH "pseudo-phonetic transcriptions of speech" (Jeremy Whistle),
the apostrophe is often at morpheme boundaries as well (as in <l'aut'c�t�>
for <l'autre c�t�> 'the other side' or <m'enfin> for <mais enfin> 'but after
all'), but it need not be, e.g. <p'tet'> for <peut-etre> 'maybe', <p'tit>
for <petit> 'small'.
 Miikka-Markus Alhonen points out that the FINNISH apostrophe marking
word-final vowel omission in poetic language (mentioned in 2. above) can
appear before a pause as well, so that he does not find it very appropriate
to speak of a _boundary_ here. Consequently, this use of the apostrophe
would have to be classified in this chapter, too.

 The GERMAN written forms <M'Gladbach> for <M�nchengladbach>, <D'dorf>
for <D�sseldorf>, <K'lautern> for <Kaiserslautern> (all place names) and
<Akk'obj.> for <Akkusativobjekt> 'direct object' (cf. Peter Gallmann,
_Graphische Elemente der geschriebenen Sprache_, T�bingen 1985, 248) are
pure abbreviations, since the words are exclusively pronounced in their full
form (not as */"klaUt6n/!). These cases should not be confused with the
Berlin street name <Ku'damm> for <Kurf�rstendamm>, which is pronounced as
/"ku:dam/, so that the apostrophe marks both elision and the morpheme
boundary before <damm>.
 HEBREW <PRWP'> (something like <PROF'>) for 'professor'. "For an
abbreviation based on more than one word, the double apostrophe is used, and
it is placed between the last two letters. These letters may not come from
different words. For example, The United States is called Arcot HaBrit in
Hebrew; the abbreviation is <ARH"B>" (Yehuda N. Falk). Note that "the
position of the apostrophe does not necessarily mark the place of omission",
as in the 'USA' example. -- This practice reminds me of Swedish
abbreviations, which can have either a full stop (one dot) at the end or a
colon (two dots) in the middle (e.g. <b:to> for <brutto> 'gross (weight)',
<n:a> for <norra> 'northern').

 After ANCIENT GREEK letters an apostrophe means that the letters have to
be read as number signs (alpha for '1', beta for '2', etc.).
 Yehuda N. Falk adds: "A similar use of the apostrophe in HEBREW is when
letters are used as numbers. A single letter is followed by a single
apostrophe, a sequence of letters representing a single number has the
double apostrophe between the last two letters. This is not always done;
when it is clear that letters are being used as numbers (e.g. in numbering
lists or on a calendar) the apostrophes are usually omitted. -- A more
recent development is the use of the (single) apostrophe as a diacritic
essentially equivalent to the hachek. For the sounds [tS], [dZ] and [Z],
Modern Hebrew orthography uses <c'> <g'> <z'>."
 In CZECH and SLOVAK, the apostrophe is used to replace the "hacek" (or
"caron") above some letters, which denotes the palatal pronunciation of
consonants, namely <d'> for palatal /d'/ (where the upper-case letter is
<D+hacek>), <t'> for palatal /t'/ (where the upper-case letter is
<T+hacek>), and (in Slovak only) <l'> and <L'> for palatal /l'/; the "hacek"
is not replaced in <n+hacek> and <N+hacek> for palatal /n'/.
 In ITALIAN, the apostrophe is often used to replace the grave accent
when words are written entirely in capitals, e.g. <CITTA'> for <citta>.
 In 19th-century FINNISH, the apostrophe was written at word boundaries
after words that cause the first consonant of the following word to be
lengthened, e.g. <vastaa' siihen> /vAstA: s:i:hen/ 'answer to it!', <kutsu'
eteen> /kutsu?: ete:n/ 'invite [him] to the front!'. Although this is
distinctive (cf. <vastaa siihen> /vAstA: si:hen/ '[he] answers to it',
<kutsu eteen> /kutsu ete:n/ 'invitation to the front'), the apostrophe is
not written any more nowadays.
 In one of the orthographies of BRETON, <c'h> is used as a grapheme for
[x] or [h] (in contrast to <ch> for /S/, as in French), e.g. <mersi braz
deoc'h> 'thanks a lot to you', <ar c'hi> 'the dog'. In another orthography,
these forms are written <deoh>, <ar hi>. Furthermore, there was a proposal
(by Steve Hewitt in "La Bretagne Lingusitique", 1986?, reported by Johannes
Heinecke) to contrast [x] and [h] by writing <c'h> for [x] and <c"h> for
[h], but this has never been done in practice.
 In SWAHILI, <ng'> is used as a grapheme for /N/ (the last sound in
"sing"), in contrast to <ng> without the apostrophe for the phoneme
combination /Ng/, e.g. <ng'ombe> (morphologically ng'o+mbe) 'cow',
<nung'uniko> (nu+ng'u+ni+ko) 'moaning'; a minimal pair is <ngambo> /Ngambo/
'a kind of tree' vs. <ng'ambo> /Nambo/ 'oversea'.
 In phonetic TRANSCRIPTION of RUSSIAN and other Slavonic languages, the
apostrophe also denotes the palatalization of consonants, e.g. </v'ek/> for
Cyrillic <BEK> 'century, age' (transliterated <vek>). This use is not
official IPA (where palatalization is indicated by a superscript <j>), but
accepted in the SAMPA adaptation of the IPA for e-mails etc.
other languages, says Mark A. Mandel), the apostrophe indicates that the
preceding consonant letter is a glottalic ejective. As symbol No. 401, the
apostrophe is also used as a diacritic denoting 'ejective' in the
 In at least one system of SHORTHAND (the Russian one), the apostrophe is
used to denote ordinal numbers, e.g. while <2> is read as _dva_ 'two', <2'>
is read as _vtoroj_ 'second'.

 In Arabic loanwords in TURKISH the apostrophe marks the glottal stop,
e.g. <mer'i> 'valid, in force'.
 Probably this use in Turkish is related to the use of the apostrophe in
the TRANSLITERATION of ARABIC and HEBREW, where it also denotes the glottal
stop, corresponding to the consonant letters 'alif and aleph respectively. I
suppose that the apostrophe in Arabic transliteration is older than its use
in Turkish roman orthography, which was introduced only in 1929. (As Mark A.
Mandel points out, the apostrophe is also used in this sense in the StarTrek
language Klingon, but that is probably patterned on the Turkish, Arabic and
Hebrew examples.)
the apostrophe is used to represent the so-called 'soft sign', a letter of
the Cyrillic alphabet basically indicating the palatalization of a preceding
consonant or representing /j/. The 'hard sign' in Russian is transliterated
either as a double apostrophe <"> or as a hyphen <->.
 The apostrophe also denotes a glottal stop in the orthography of
HAWAI'IAN (cf. the Hawai'ian name of the state itself <Hawai'i>) and
 The UKRAINIAN apostrophe described in 1. above, which indicates /j/, is
also used in foreign words where there is no morpheme boundary, e.g.
<p'jesa> for /pjesa/ '(stage) play'. If one counts these cases, one has to
classify the apostrophe as a quasi-letter representing /j/, or else one has
to view <p> in the word <p'jesa> as an orthographic pseudo-prefix.

Another function of the apostrophe, as Miikka-Markus Alhonen points out
correctly, is its use as a single QUOTATION MARK. I am not sure whether the
quotation marks are in any way related to the apostrophe or whether their
graphical identity is just incidental. What is different in quotation marks
is that they appear in pairs (and except in Finnish and on typewriters,
opening and closing quotes usually look different, a feature they have in
common with parantheses), and quotation marks have many different forms in
different languages (guillemets <<...>> and >>...<<, opening quotes at the
bottom ,,..." etc.), whereas the apostrophe looks more or less the same in
all the languages I know. (In Hebrew it is mirrored because Hebrew is
written from right to left.)

A very special case of apostrophe use, which I am not sure how to classify,
has shown up in an experiment conducted by Christine Haunz. She examined the
perception and reproduction of unfamiliar (in this case: Russian) consonant
clusters by English native speakers (who do not have any knowledge of
Slavonic languages). In one of her experiments the candidates had to
transcribe an aural input, and many inserted vowel letters into the
consonant clusters, just as in other experiments they pronounced the
clusters with some kind of vowel (mostly shwa) when they had to repeat the
input. Two of these clusters were word-initial /tv/ and /dv/, and one of the
test candidates transcribed these as <t'v> and <d'v>, respectively. What
does that mean? Did the test person hear some vowel between the consonants
but did not know which, and did (s)he draw a parallel to forms like <isn't>,
which are sometimes pronounced with a shwa as well (i.e. ["Iznt])? Or did
(s)he hear nothing between the consonants but knew that a consonant cluster
like this cannot occur in English, so there must be "something missing"
inbetween, or did (s)he just "avoid" this unfamiliar combination by
separating the two letters with an apostrophe (cf. the French example
<N'Diaye> in 1. above)?

Please let me pay special thanks to all those who discussed this subject
with me in long e-mail conversations (especially Miikka-Markus Alhonen,
Peter T. Daniels, Johannes Heinecke, Wolf Peter Klein, Donald F. Reindl,
Charley Rowe, and Nino Vessella). With counter-arguments against my thesis
they have helped me refine my argument. Some of my weaker arguments, which
have been refuted successfully, do not appear in this summary any more
thanks to them.

For those who are interested: I will now go through all this material again
and use it for a paper to be read in September at the Third International
Workshop on Writing Systems "From Letter to Sound" in Cologne
(cf. The abstract is already
online (, and the whole
paper will be published in the conference proceedings.

Daniel Buncic
Bonn University Seminar of Slavonic Philology
Lenn�str. 1, D-53113 Bonn
Phone: +49 228 73-7203
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue