156 Giovanna Marotta
In particular, the process of l-velarization exhibits a variable degree of thick-
ness, since its constraints are twofold: lexical (frequency of the word) and pho-
netic (vowel quality; see §5.6). With reference to a gradual scale of thickness, the
picture can be summarized as follows:- the highest value has to assigned to words such as bello ‘nice’, fratello ‘brother’,
sorella ‘sister’, which are very frequent and contain a geminate lateral and
a low [æ] vowel (which is one of the strongest shibboleth of Leghornese;
Calamai 2004); - an intermediate value has to be assigned to words such as polpo ‘octopus’,
palmo ‘palm’, facoltà ‘faculty’, which are also relatively frequent and contain a
low or a mid back vowel; - the lowest value has to be assigned to words such as birillo ‘pin’, grullo ‘f o o l i s h’,
fanciullo ‘child’, which are relatively infrequent and contain high vowels.
In conclusion, thickness may be considered as a parameter for evaluating the
degree of robustness of a sociophonetic index, according to the implication in (17):(17) Sociophonetic index:
more difficult to control → more frequent → more robust → thicker- Weight
Coming back to our metaphor of solids and to the theory of solids in physical
geometry, a new analogy may be recognized: in physics, solids have mass and
weight, which are properties belonging to them as objects embedded into the
physical world. The abstract conception of solids does not entail the property of
weight, which becomes indeed necessary where a solid is inserted in a physical
environment. In a similar way, sociophonetic indexes do not have weight per
se, while they receive a different weight in their usage in everyday speech. The
linguistic community assigns the weight. The values of this parameter can be
different according to the social evaluation of a given phonetic cue.
In some way, weight appears to be the strongest sociolinguistic property
in the proposed set of parameters, because it refers to the social value given to
the sociophonetic indexes by the speakers of a linguistic community. What I call
weight is basically equivalent to the term ‘prestige’, traditionally employed in
sociolinguistics (Trudgill 1972; Labov 2001: 60, 196 ff.). As is well known, the
sociolinguistic prestige of a variable is directly proportional to its use in more
formal styles. With respect to weight, the attributes of heavy and light may eas-
ily correspond to covert-prestige and overt-prestige, respectively. In our opinion,