There are also situations which are going in reverse directions : added consonants which do not carry any semantic value and only added for phonologic reasons to force the separation of syllables, and which are definitely not null consonnants.
Think about the -z- and -t- particles used in French between some short words (notably after conjugated verbs and before one of the pronouns "y" and "en", limited otherwise to only a single vowel sound which need to be stressed separately): "Comment va-t-il ?", "Va-z'y !"... You cannot understanrd their role if you just look at the mening of each word, and how verbs are conjugated. A hiatus could be used (with a "null" consonnant, or a glotal stop), but the oral language developed these heard particles instead in some common situations (but most tables showing how to conjugate verbs forget speaking about when these contextual particles are not just possible or desired, but in fact now required by the modern language...). These particles initially originate from the early formation of "liaisons" (which are traces used only in the oral language, of final consonnants that are have paridly become mute, but were kept by tradition in the written form as they may form liaisons in some situations). Other languages use phonologic particles, but more frequently they use mutations of initial consonants, or elision of some initials or finals, or will use contractions of words (contractions also exist in French, like in Italian, and even in English for the negation of auxialiry verbs). Not all of them are transcribed in the written form of the language. 2013/9/20 Richard Wordingham <[email protected]> > On Fri, 20 Sep 2013 02:34:45 -0700 > Stephan Stiller <[email protected]> wrote: > > > On 9/19/2013 2:35 AM, Stephan Stiller wrote: > > >> As far as I am aware, a proper 'null consonant' has only arisen > > >> when it actually represents a glottal stop. > > > > There's ㅇ in hangeul ... > > > And I'll add that the letter "h" in various Romance languages... > > I was thinking of abugidas, or, more precisely, Indic scripts. > Consonant sounds fading out should be another route, but I can't think > of any examples with Indic scripts. > > Richard. > > >
