150 Giovanna Marotta
sibilant gives rise to an affricate and Italian has a dental voiceless affricate as a
phoneme, which can occur even in the context triggering /s/ affrication (e.g., alza
‘(he) raises’; danza ‘danse’).
Therefore, a ‘near-merger’ of the two phonological categories arises,^11 because
when a Tuscan speaker-listener hears a phonetic string like [-nts-], [-rts-], [-lts-],
he/she cannot know whether there is a /s/ or a /ts/ phonemic category, unless he
knows the meaning of the word and how it is written. The process of s-affrication
is so pervasive that nowadays children have problems at school in writing even
common words such as penso ‘(I) think’, salsa ‘sauce’, borsa ‘ b a g’, polso ‘ w r i s t’,
perso ‘lost’, which are often written as <penzo, salza, borza, perzo, polzo, perzo>.^12
Note the fact that the process may give rise to the neutralization of the phonemic
contrast between /s/ and /ts/. For instance, the two words salsa ‘sauce’ /ˈsalsa/ and
Salza /ˈsaltsa/ ‘the name of a Pisan cake shop’ collapse in a unique phonetic form,
i.e. [ˈsaltsa], in Tuscan pronunciation. Similarly, the distance between some lex-
emes may decrease. Consider for instance orso ‘bear’, SI /ˈorso/ and orzo ‘barley’,
SI /ˈɔrdzo/: the two words are pronounced [ˈortso] and [ˈɔrdzo], respectively, both
with an affricate, though with a different voicing value.
The acoustical analysis by Turchi & Gili Fivela (2004) has shown that stop
closure is short and less strong in the case of /s/ becoming an affricate with respect
to an underlying /ts/. As a matter of fact, the fine phonetic detail (in the sense of
Local 2003) of a few milliseconds of difference in the stop closure can easily be
recognized at the articulatory and acoustic levels, without having any clear cor-
respondence at the perception level: both underlying /ts/ and the affricate stem-
ming from post-sonorant /s/ are actually perceived as the same sound, with the
consequent suspension of a phonemic contrast.
- The term ‘near-merger’, coined by Labov (1966), has been deeply discussed by Labov himself
(1994: 349–370). He underlines the disbelief of the phenomenon by linguists and phoneticians,
despite the wide range of data supporting the idea of partial neutralizations or near-mergers.
From the theoretical point of view, it is quite obvious that traditional phonology cannot recog-
nize the notion of near-merger, inasmuch as it challenges the principle of binary opposition as
well as the symmetry between production and perception. - With respect to the affrication process of /s/, a very nice example may be quoted here: in
the object of an e-mail message written on November 2010 by a University caretaker of the
Department of Linguistics of Pisa University, the following phrase was written: sospenzione
della didattica in data 30 Novembre 2010 (i.e. ‘Lessons are suspended on November 30th 2010’),
instead of sospensione. Who wrote the message was unable to identify the phoneme corre-
sponding to the segment [ts] in a post-consonantal context: is it /s/, to be written as, or
/ ts/, to be written as? Nowadays, this is the Hamlet’s question for many Tuscan speakers,
listeners and writers.