Introduction

The investigation of diminutive, augmentative, ameliorative and pejorative morphemes started with Scalise (1984), who also coined the term evaluative morphology to refer to the encoding of the corresponding diminutive, augmentative, endearment and pejorative features through morphological operations, mainly consisting in the addition of dedicated evaluative suffixes.

Generally speaking, evaluative features relate to the semantic primitives BIG, SMALL, GOOD, BAD (Wierzbicka 1989; Grandi 2002) and express the objective alteration of the size of the referent (diminutive and augmentative), or its subjective evaluation (endearment and pejorative) with respect to a standard. To be more specific, diminutive and augmentative features convey a descriptive or quantitative evaluation referring to the denotation of the referent. On the other hand, endearment and pejorative features express a qualitative evaluation, or a judgment, with respect to the connotation of the referent.

The pioneering study on Italian elaborated by Scalise (1984) started a tradition of cross-linguistic researches, whose main results have been collected in Grandi and Körtvélyessy (2015). These studies have shown the universality of evaluative features, though their encoding is language-specific. Indeed, evaluative features can either involve dedicated morphemes, or other linguistic strategies such as tonal variations, vocalic patterns, or even syntactic constructions involving adjectives, depending on the morphological type of the language. Therefore, the domain of evaluation involves more than just morphology, and the linguistic operations carrying evaluative features are better defined as evaluative constructions rather than evaluative morphology.1 Despite this heterogeneity, in general, evaluative constructions consist of an element encoding the evaluative feature (evaluative marker) and of an additional one, usually the stem or an independent word, encoding the standard referent of the evaluation.

The domain of evaluative constructions is not only complex with respect to the linguistic strategies involved to express evaluative features. It also displays a complex system of meanings combining semantic and pragmatic factors. In the most common case, evaluative markers carry the meanings of the semantic primitives, namely ‘big’, ‘small’, ‘good’, ‘bad’. However, this is not the whole story. Evaluative markers can convey a number of emotional overtones together with the canonical meanings of the primitives stated above. In other words, evaluative constructions constitute a very creative domain (see Dressler and Merlini Barbaresi 1994 for an overview). For instance, the same language can employ a sequence of different morphemes to convey that something is lovely and small, or it can express different emotional overtones while employing one single morpheme. Italian is a very good example: the word casina contains the diminutive suffix -in(a), which usually denotes the small dimension of the object defined by the base, here cas- ‘house’, and conveys the meaning ‘small house’. Nevertheless, in some contexts the diminutive marker can also encode the affection of the speaker towards the object of the evaluation, thus functioning as an endearment marker. In so doing, the morpheme -in(o/a) in Italian does not necessarily refer to a small object, but rather conveys the meaning ‘lovely house’ or ‘lovely little house’. This is to show how the meanings and functions of evaluative markers across languages heavily depend on both semantic and socio-cultural or pragmatic factors. The majority of the literature on evaluative constructions concentrates on these aspects and investigates both language-specific alternations as well as general properties holding cross-linguistically. Investigations carried out on hundreds of spoken languages show that evaluative constructions are typologically similar across different languages. Among the studies exploring the semantic and pragmatic aspects of evaluative markers, we find some authors who claim that diminutive and augmentative features are encoded in language because the concept they convey, namely size, is bodily perceived by humans and thus can be expressed through language (Prieto 2005, 2015).

From the issues considered so far, it is clear that evaluative constructions, despite their apparent simplicity, need to be investigated against different theoretical backgrounds in order to be completely understood. I argue that we should combine the generative, cognitive and typological tradition. I am aware of the fact that these approaches are very different to some extent, however, I think that evaluative constructions can constitute the point of encounter. Indeed, the fact that all the languages investigated so far are endowed with some elements encoding the evaluative features is evidence for the universality of evaluative features in generative terms. This innate endowment is further supported by the fact that humans are ‘wired’ (in cognitive terms) to encode concepts such as size, since they can perceive it. Nevertheless, this universality could not be claimed if typological studies investigating several languages did not exist. It follows straightforwardly that we can only fully understand evaluative constructions by taking into account all of these perspectives. However, one further piece must be added to the puzzle to really gain completeness: sign languages.

Sign languages are fully-fledged natural languages developed across the Deaf2 communities around the world. They are visual-manual (also called visual-gestural or visual-spatial) languages conveyed through hands, body movements and facial expressions, and visually perceived. The first studies investigating linguistic and psycholinguistic aspects of sign languages date back to the early 1960s. Since then, a still growing body of literature has developed to account for their structure and grammar, which conform to the principles of discreteness and recursiveness (Chomsky 1957; Hockett 1960) defined for spoken languages, despite the different modality employed. Therefore, a linguistic property can really be considered universal only if it holds for the languages in both modalities, spoken and visual-manual. Cross-modality researches have been, and still are, crucial in that they allow us to confirm existing generalisations, but also to improve the knowledge of sign language-specific characteristics which distinguish them from spoken languages.

This is the reason why it is necessary to include sign languages also within studies exploring evaluative constructions. By investigating the display of evaluative features in sign languages we can: (i) confirm the universality of evaluative features; (ii) improve the knowledge we have of the phenomenon while adding sign-language specific strategies for encoding evaluative features; (iii) confirm the assumption that some concepts are expressed in the language because they are bodily perceived, as is the case for the concept of size. Being mainly manually conveyed, sign languages can provide evidence for the embodiment of concepts through language. The general picture provided by Petitta, Di Renzo, and Chiari (2015) is a first step toward the understanding of evaluative morphology in sign languages. In this book, I focus on the display of evaluative features in Italian Sign Language (LIS).

LIS is the language used by the Deaf community living in Italy. Linguistic investigation on LIS started in the late 1970s at the National Research Council (CNR) in Rome thanks to Virginia Volterra and a group of scholars including Elena Pizzuto, Elena Radutzky, Serena Corazza, among others. The pioneering researches aimed at showing that LIS was not a basic gestural system, but rather a fully-fledged natural language endowed with a specific grammar as all natural languages are (Volterra 2004). In so doing, they studied the structure of LIS grammar and the processes of acquisition in a sign language. These studies initiated a long tradition of linguistic and psycholinguistic research, whose main achievement has been to recognise the linguistic status of LIS both across the Deaf and the hearing communities living in Italy. The first researches mainly explored the syntax of LIS and its complex classifier system. As far as the morphology of LIS is concerned, there are studies concerning nominal (Pizzuto and Corazza 1996) and verbal inflection (Corazza 2000), but derivational processes are less investigated.

Therefore, the aim of my research is threefold. First, I aim at providing a detailed description of the encoding of evaluative features in LIS through the analysis of both corpus data and elicited data. I will do so by considering the markers involved, the relation with the lexical adjectives corresponding to the semantic primitives, the nature of the detected morphological processes (sequential or simultaneous), and the constraints allowing some strategies rather than others. Second, I aim at improving typological studies with data from a sign language by comparing the characteristics of the evaluative strategies detected in LIS to the properties usually associated with evaluative constructions in spoken languages. Third, I will investigate the applicability of formal derivations to account for LIS evaluative constructions.

The book is organised as follows. It comprises three parts devoted to the theoretical backgrounds, the methodological...