My research interests mainly concern phonetics and phonology, with
specific attention to speech production and perception of both
intonation and segments in L1, L2 and pathological speech; recently, I
also explored the use of Natural Language Processing techniques to
reach more application-oriented goals, of course in team work.
I believe intonation represents a very interesting prosodic trait to be
investigated from a linguistic point of view. Starting with my PhD
studies on the phonetics and phonology of intonation in Pisa Italian, I
marveled at observing the amount of linguistic information conveyed
by the speaker modulation of the fundamental frequency of her/his
voice and the related prosodic changes. Later I’ve got also interested in
the inter and intra-system variation that is observed when investigating
the phonology of intonation. In this respect, I am currently devoting
specific attention to various types of contact that may affect
intonational systems (variety-variety, dialect-variety, L1-L2) and to
issues concerning the tune-text relation.
I think adopting different methods may be useful to disentangle
complex linguistic phenomena. Thus, acoustic analyses often represent

only the starting point of the experimental protocols I plan, where
perception (investigated behaviorally and, in certain cases,
neurophysiologically) and speech articulation are also studied. Speech
perception investigations are crucial for understanding the relevance of
acoustic events within the linguistic system. In my research, perception
studies have been decisive for investigating intonation events as well as
segmental aspects of L1 and L2 speech, also in their interaction with
prosodic structures such as the syllable.
I’ve also found the investigation of the geminate and singleton
consonants in Italian particularly interesting, as it shows the tight
relation between segmental and prosodic aspects and structures.
Further, the work on geminates first allowed me to step into the world
of speech gestures and motor control. This turned out to be crucial in
studying effectively the organization of the phonetic-phonological
system in different L1s, as well as its reorganization in L2 acquisition
and in pathological speech.
As for the latter, I have been working on motor speech disorders such
as dysarthria, in particular hypokinetic dysarthria that usually develops
in Parkinson’s Disease. The analysis of speech kinematics in dysarthric
speech, in particular by adopting the ElectoMagnetic Articulography
(EMA) or the Ultrasound Tongue Imaging systems (UTI), is fruitful in
understanding the impact of the disease in speech articulation, offering
details on both the characteristics of speech gestures and the
articulatory strategies adopted to overcome the difficulties caused by
the disease.
Finally, in some works done in collaboration with Natural Language
Processing experts, I checked the use of deep linguistic models, and
natural language processing techniques in general, for automatically
classifying and evaluating speech produced by healthy and pathological
speakers, such as dysarthric patients with Parkinson Disease or subjects
affected by dementia.