Evaluatives in Morphology
Summary and Keywords
Evaluative morphology is a field of linguistic studies that deals with the formation of diminutives, augmentatives, pejoratives, and amelioratives. Actually, evaluative constructions cross the boundaries of morphology, and are sometimes realized by formal strategies that cannot be numbered among word formation processes. Nevertheless, morphology plays a dominant role in the formation of evaluatives. The first attempt to draw an exhaustive account of this set of complex forms is found in the 1984 work Generative Morphology, by Sergio Scalise, who made the hypothesis that evaluatives represent a separate block of rules between inflection and derivation. This hypothesis is based on the fact that evaluatives show some properties that are derivational, others that are inflectional, and some specific properties that are neither derivational nor inflectional. After Scalise’s proposal, almost all scholars have tried to answer the question concerning the place of evaluative rules within the morphological component. What data reveal is that, in a cross-linguistic perspective, evaluatives display a uniform behavior from a semantic and functional point of view, but exhibit a wide range of formal properties. In other words, functional identity does not imply formal identity; consequently, we can expect that constructions performing the same function display different formal properties in different languages. So, if evaluatives are undoubtedly derivational in most Indo-European languages (even if they cannot be considered a typical example of derivation), they are certainly quite close to inflection in some Bantu languages. This means that the question about the place of evaluatives within the morphological component probably is not as crucial as scholars have thought, and that other issues, sometimes neglected in the literature, deserve the same attention. Among them, the role of pragmatics in the description of evaluatives is no doubt central. According to Dressler and Merlini Barbaresi, in their 1994 work, Morphopragmatics: Diminutives and Intensifiers in Italian, German and Other Languages, evaluative constructions are the more typical instantiation of morphopragmatics, which is “defined as the area of general pragmatic meanings of morphological rules, that is of the regular pragmatic effects produced when moving from the input to the output of a morphological rule.” Evaluatives include “a pragmatic variable which cannot be suppressed in the description of [their] meaning.” Another central issue in studies on evaluative morphology is the wide set of semantic nuances that usually accompany diminutives, augmentatives, pejoratives, and amelioratives. For example, a diminutive form can occasionally assume a value that is attenuative, singulative, partitive, appreciative, affectionate, etc. This cluster of semantic values has often increased the idea that evaluatives are irregular in nature and that they irremediably avoid any generalization. Dan Jurafsky showed, in 1996, that these different meanings are often the outcome of regular and cross-linguistically recurrent semantic processes, both in a synchronic and in a diachronic perspective.
Within the field of linguistic studies, the labels ‘evaluation’ or, more often, ‘evaluative morphology’ (alternative labels usually attested in the literature are appreciative, expressive, alterative morphology) are used to indicate a wide and varied set of linguistic constructions performing different semantic functions, which can be summarized as in the following list:
The formal strategies more frequently used to express these functions are
These strategies can be combined, so multiple evaluation constructions (that is, constructions where two or more different strategies occur at the same time) are attested:
As can be seen, the domain of evaluatives crosses the boundaries of morphology. Consequently, it can be stated that morphology is, cross-linguistically, the most frequent realization of evaluation, but a full identification of evaluation with morphology is incorrect. That said, this article will focus on morphological evaluatives.
The vast majority of examples presented and discussed in this contribution are drawn from the recently published Edinburgh Handbook of Evaluative Morphology, edited by Nicola Grandi and Livia Körtvélyessy (2015b).
While evaluatives have long been identified as a specific research area, the first attempt to give them a specific account is Scalise (1984, pp. 132–133; and see Scalise, 1994, pp. 264–266). Even if the titles of the chapters devoted to this issue are ‘Evaluative Suffixes’ in Scalise, 1984, and “Suffissi valutativi ‘evaluative suffixes’” in Scalise, 1994, the author’s attention is exclusively paid to the Italian diminutive suffix –ino; in his 1994 book, Scalise asserts that the same generalizations hold for the other groups of suffixes, pejoratives, and augmentatives above all. Scalise recognizes six properties that seem to be characteristic of evaluative morphology:
1. The syntactic category of the base they are attached to does not change (tavolo ‘table.N’ – tavol-ino ‘house-DIM.N, small table’; giallo ‘yellow.ADJ’ – giall-ino ‘yellow-DIM.ADJ, yellowish’; bene ‘well.ADV’ – ben-ino ‘well-DIM.ADV, so so’).
2. The syntactic features or the subcategorization frame of the base are not changed (letto ‘bed.N[concrete]’ – lettino ‘little bed.N[concrete]’; idea ‘idea.N[abstract]’ – ideuzza ‘little idea.DIM.N[abstract]’; giocare ‘play.V[transitive]’ – giocherellare ‘play around.V[transitive]’).
3. The semantics of the base word is changed (macchina ‘car’ – macchinina ‘toy car’; lume ‘lamp’ – lumino ‘little lamp’).
4. Evaluative suffixes allow the consecutive application of more than one rule of the same type: the result of every application is an actual word (cf. fuoco – fuocherello / fuocherellino ‘fire – little fire – nice little fire’; uomo – omaccio – omaccione ‘man – bad man – big bad man’).
5. Evaluative suffixes are always external with respect to other derivational suffixes and internal with respect to inflectional morphemes (e.g. contrabbandierucoli ‘small time smugglers’ < compound contrabbando [‘contraband’] + derivational suffix [-iere ‘smuggler’] + evaluative suffix [-ucol(o) diminutive / pejorative] + Inflectional morpheme [-i masculine, plural]).
6. Evaluative suffixes can trigger specific readjustment rules; some of them are fully predictable, others are idiosyncratic (an example of the first kind of rules is the insertion of [tʃ] between the base and the diminutive suffixes –ino and –ello if the base ends in [‘one’]/ [‘ɔne’]; an example of an unpredictable readjustment rule is the insertion of a consonant before the diminutive suffix with bases ending in a stressed vowel: caffè ‘coffee’ – caffettino / cafferino; città ‘town’ – cittadina, etc.).
According to Scalise, the properties (1) and (2) are typically inflectional; the properties (3) and (4) are typically derivational; finally the properties (5) and (6) seem to be typical of evaluative affixes and distinguish them both from inflection and derivation. Due to this situation, it is not possible to include evaluative morphology in the realm of either derivation or inflection. Consequently, Scalise (1984, p. 144) proposes the idea of a third morphology: “In a level ordered morphology such as the one assumed here, this situation can be handled easily by ordering a separate block of "Evaluative Rules" (ER's) after the derivational (Word Formation Rule) WFR's and before the Inflectional Rules (IR’s)”:
Even if Scalise’s criteria were set specifically for Italian, and even if his statement about the possibility of extending his generalizations to other classes of evaluative affixes has been contradicted by data, his studies had the great merit of making clear that evaluative morphology deserves a specific account within the domain of word formation processes, and that it is not a mere peripheral, marginal, or irregular set of forms: from here on in, evaluative morphology has become a more and more central issue in morphological investigations.
But this had also a side effect: for many years the issue concerning the place of evaluative morphology with respect to inflection and derivation monopolized the debate, putting aside other crucial points. This was due to the fact that some details have often been ignored. First, all non-affixal evaluatives have been neglected. Second, it has often been assumed that functional identity implies formal identity; so, since evaluatives play the same function, they must display the same formal behavior in a cross-linguistic perspective. Of course, it is not true: evaluatives exhibit a wide range of formal properties even if they perform the same function (so, for example, Italian diminutives do not trigger any agreement, while Gikuyu diminutives do, etc.). Third, the role of pragmatics has never been seriously taken into account, at least until Dressler and Merlini Barbaresi (1994). According to them, evaluatives must be dealt with primarily in terms of pragmatics or, better, morphopragmatics. Morphopragmatics is “defined as the area of general pragmatic meanings of morphological rules, that is of the regular pragmatic effects produced when moving from the input to the output of a morphological rule” (Dressler & Merlini Barbaresi, 1994, p. 51). In this picture, evaluatives are considered the more typical instantiation of morphopragmatics. Contrary to a widely accepted view, according to which “pragmatics is usually seen as an antithesis to grammar” (1994, p. 50), in the opinion of Dressler and Merlini Barbaresi (1994, p. 55), evaluative word formation processes include “a pragmatic variable which cannot be suppressed in the description of [their] meaning.” This variable, if present, is not unpredictable (for diminutives, it is the feature [non-serious]; Dressler & Merlini Barbaresi, 1994, p. 144): some morphological processes have regular pragmatic effects; this means that some evaluative constructions are the regular formal correlate of extralinguistic and contextual constraints; in other words, as claimed by Merlini Barbaresi (2015, p. 32), to whom the reader may refer for more details, “some productive morphological operations are autonomously responsible for effects that cannot be exhaustively explained solely by their semantics plus generic contextual conditioning.” For a pragmatic view of evaluatives, see also Schneider (2013).
The pragmatic role played by most evaluatives has undoubtedly contributed to the stereotype that their behavior is largely irregular. Another situation that has made this prejudice stronger is “the fact that the meaning of evaluative constructions usually displays a multiplicity of senses.” Consider, for example, the case of Italian suffix –ino, which can occur in complex words expressing diminution in size (librino ‘small book’ < libro ‘book’), endearment (maritino ‘dear husband’ < marito ‘husband’), contempt (dottorino ‘untrustworthy doctor’ < dottore ‘doctor’), singulative (zuccherino ‘sugar lump’ < zucchero ‘sugar’); it is also used in the formation of instrumental nouns (frullino ‘(small) mixer’ < frullare ‘to mix’; cf. frullatore ‘mixer’), of ethnic adjectives (tunisino ‘Tunisian’), etc. As stated by Haspelmath (2003, p. 211), these kinds of morphemes “have more abstract and general meaning and are thus more apt to be used in multiple ways than content words.” The cluster of meanings, often associated to evaluative affixes, contributed to accentuate “the idea of their anomalous nature” (Grandi & Körtvélyessy, 2015a, p. 5).
According to Simon and Wiese (2011, p. 21), there is a correlation between the general (or underdetermined) meaning of a linguistic construction and the development of seemingly exceptional patterns. Among other cases, they discuss just the behavior of diminutives, asserting that “their erratic behaviour turns out to be more systematic when viewed from a morphopragmatic perspective” (2011, p. 22), in this way confirming the point of view expressed by Dressler and Merlini Barbaresi (1994) and Merlini Barbaresi (2015). In other words, “the possibility of directly involving pragmatic aspects in morphology” (Simon and Wiese, 2011, p. 22) explains some of the seemingly unusual properties of diminutives, augmentatives, etc.
If we turn to semantics, the constellation of meanings that characterizes evaluatives is far from chaotic and unpredictable. Most meaning associations have been demonstrated to be the outcome of productive and cross-linguistically recurrent semantic processes (metaphor, inference, etc.). The first attempt to draw a consistent picture of the polysemy of evaluatives is Jurafsky (1993 and 1996), whose analysis is limited to diminutives. Jurafsky organizes all semantic and pragmatic values of diminutives in a radial category:
According to Jurafksy (1996, p. 542), “a radial category consists of a central sense of prototype together with conceptual extensions, represented by a network of nodes and links. Nodes represent prototypes of senses, while links represent metaphorical extensions, image-schematic transfer, transfers to different domains, or inferences.” A similar approach is that of Prieto (2005, 2015, pp. 27, 28), who extends the analysis to augmentatives:
According to this model, evaluatives are characterized by a structured polysemy.
According to Jurafsky (1996, p. 542), a radial category can be used both in synchronic and historical description of evaluatives: “Thus, when interpreted as a synchronic object, the radial category describes the motivated relations between senses of a polysemous category. When interpreted as a historical object, the radial category captures the generalizations of various mechanisms of semantic change.” Mutz (2015) demonstrates that even if the diachronic processes that take place in the development of evaluatives can be represented by a radial category, it is not possible to use a unique model to represent both synchronic and historical variation.
A recurrent feature in the literature on evaluatives is that most studies lack an explicit definition of ‘evaluation’: as a consequence, we cannot be sure that scholars are investigating the same data and are sharing the same conception of evaluation. Therefore, their analysis of data and their conclusions can hardly be compared. In this section, I’ll try to draw an explicit definition of ‘evaluative construction’ and make clear the parameters governing the selection of relevant data. I will refer to Grandi (2002, 2005) and Grandi and Körtvélyessy (2015b).
As stated by Haspelmath (2010, p. 665), “typologists have often observed that crosslinguistic comparison of morphosyntactic patterns cannot be based on formal patterns (because these are too diverse), but has to be based on universal conceptual-semantic concepts.” He adds, “we need to distinguish carefully between descriptive categories, that is, categories of particular languages, and comparative concepts, which are used for crosslinguistic comparison […]. Descriptive formal categories cannot be equated across languages because the criteria for category assignment are different from language to language” (2010, p. 663).
In this picture, we can assert that Scalise’s proposal of a third morphology is based largely on descriptive categories—in other words on formal patterns that have been revealed to be specific to Italian and genetically contiguous languages; but “for morphosyntactic comparison to be possible, we must hold the meaning constant” (Haspelmath, 2007, p. 127).
As claimed by Grandi and Körtvélyessy (2015a, p. 9), “it is crucial to assume that different linguistic items can be considered part of the same class, if they perform the same function, as this provides a cross-linguistically valid definition of evaluative morphology. Performing the same function does not assure formal identity and is not, if considered in isolation, a sufficient condition to establish the membership of a class, despite being a necessary condition. Of course, semantic-functional features should have formal correlates, but the very nature of these correlates falls into the set of language-specific facts. Failure to recognize the asymmetry between these two levels has often produced weak generalizations, leading some scholars to generalize facts that, on the contrary, are mere tendencies in specific languages or groups of languages (cf., for example, the almost exclusive focus on the issue of the place of evaluative morphology with respect to inflection and derivation, and the consequent oblivion of non-affixal evaluative constructions: tonal variations, compounds, reduplications, etc.).”
As demonstrated in Grandi (2002, pp. 31–34), the semantic functions listed in (1) can be reorganized in two classes according to the nature of the ‘evaluation’ performed: some functions express a descriptive (or quantitative) evaluation; others express a qualitative evaluation:
In other words, an item (be it an object, a person, an action, etc.) can be evaluated according to its objective properties (dimension, shape, etc.) and to the speaker’s feelings towards it, on the basis of subjective, individual criteria. Grandi and Körtvélyessy (2015a, p. 10) state that “a descriptive evaluation is not influenced by personal feelings or opinions, but it relies on real and objective properties of an item. In both cases, a deviation from a standard or from a default value may be observed: Slovak domisko ‘large house’ derives from dom ‘house,’ and it conveys a quantitative/descriptive deviation from the default size of houses … The standard or default value (or norm, in Schneider, 2013) is culturally and/or socially determined (it is a mental representation, shared by the members of the same community). In quantitative/descriptive evaluation, there is an objective, observable, evident deviation. In qualitative evaluation, the speaker perceives or feels a deviation. For example, if we compare a Dachshund and a Great Dane, we can call the former cagnolino (Italian, ‘small dog’) and the latter cagnone (Italian, ‘big dog’), since they are objectively smaller and bigger than a standard dog: nobody would contradict these statements, since this evaluation is not determined by the subjective perception of the speaker. This kind of evaluation, which is most often based on dimensional parameters, does not depend on subjective and personal judgments … Nevertheless, someone could call his/her Great Dane cagnolino, expressing his/her affection to it. In this case, different speakers can evaluate the same object (or action, person, etc.) in different ways … While descriptive evaluation usually relies on real and permanent characteristics of an item, qualitative evaluation is often based on temporary and variable parameters or situations: the first kind of evaluation is a description; the second one is a sort of interpretation.”
The same semantic functions can be grouped following a different parameter: evaluation can express an increase or decrease of a property or of a feeling, in both the quantitative and the qualitative perspective. These situations can be represented by a shift towards the positive (+) end or towards the negative (−) end of the axis corresponding to the semantic scale according to which the evaluation takes place:
If we combine (7) and (8), we obtain the following picture:
A way to obtain a higher degree of abstraction is to replace the semantic functions with two pairs of semantic primitives: SMALL and BIG are representative of the descriptive side of evaluation, while GOOD and BAD represent its qualitative side (Wierzbicka, 1996):
In this picture, we can give the following definition of the most frequent evaluative constructions:
• Prototypical diminutives indicate a shift towards the negative end on the descriptive axis
• Prototypical augmentatives indicate a shift towards the positive end on the descriptive axis
• Prototypical pejoratives indicate a shift towards the negative end on the qualitative axis
• Prototypical amelioratives indicate a shift towards the positive end on the qualitative axis.
(Grandi & Körtvélyessy, 2015a, p. 12)
This pattern of variation identifies the typical position of an evaluative construction. But, it is well known that an evaluative construction can display an atypical behavior, for example when the two perspectives merge. So, a single construction can show both a diminutive and an ameliorative value, as in Italian (gattino ‘kitten, baby cat’), or both a diminutive and a pejorative value (as in Italian dottorino ‘young and unexperienced doctor’). Different functions can combine in many ways, with a single restriction: contradictory values cannot be expressed by the same construction. So for example, we will not find a construction that can be interpreted as diminutive and augmentative at the same time:
Consistent with these premises, I adopt here the definition of evaluation proposed by Grandi (2002, p. 52). In short, a construction can be defined as evaluative if it satisfies two conditions, one relating to semantics and the other to the formal level. The first condition indicates that a linguistic construction can be defined as evaluative if it has the function of assigning to a concept a value, different from that of the ‘standard’ or the ‘norm’ (within the semantic scale to which it is part of), without resorting to any parameters of reference external to the concept itself. The second condition indicates that an evaluative construction must include at least the explicit expression of the standard (by means of a linguistic form that is lexically autonomous and is recognized by the speakers of the language as an actual word) and an evaluative mark (a linguistic element that expresses at least one of the semantic values traditionally classed as evaluative: BIG, SMALL, GOOD, BAD). So, in Italian gattino ‘kitten, dear little cat,’ the norm is expressed by the lexical morpheme gatt- (which occurs in masc. gatto and fem. gatta); the evaluative mark is the suffix –ino. In Apma biri biri ‘very small,’ the norm is expressed by the word biri; the evaluative mark is the full reduplication of this word.
4. Evaluatives Between Inflection and Derivation
The second part of this article deals with a comparison cross-linguistically between the most frequent properties of affixal evaluatives on one side and the typical properties of inflection and derivation on the other. Evaluative affixes (and suffixes in particular) are the linguistic strategy largely prevalent among the languages of the world for the realization of the four evaluative semantic values (BIG, SMALL, GOOD and BAD) described above (cf. Štekauer, 2015). In the literature, evaluative affixes are usually placed in a ‘grey area’ between derivation and inflection, which seems to avoid all regularities and therefore prohibits any generalization. Beard (1981. p. 180) places evaluative suffixes “somewhere between lexical and purely inflectional forms.” According to Szymanek (1988, pp. 106–109), evaluative suffixes occupy an “expressive periphery” of derivation, and they do not correspond to any cognitive category. Similaly, Carstairs-McCarthy (1992, p. 107) states that evaluative suffixes belong to “expressive morphology,” which is different from derivational morphology and exceptional with respect to its principles. Finally, as already said above, Scalise (1984, 1994) makes the hypothesis of a “third morphology,” whose place is between derivation and inflection.
Data reveal that the actual situation is, in fact, far less complicated and chaotic than it seems, provided that evaluation, even when presenting homogeneity on a semantic level, is substantially a ‘cross-cutting’ linguistic operation, as it presents different formal realizations, even within the same language. In this sense, it is not advantageous to try to establish a universal collocation of evaluative affixes within the morphological component: these, although sharing a series of formal and semantic properties, could be derivational in some languages (e.g., Indo-European languages) or close to inflexional in others (e.g., Bantu languages). If this perspective of investigation is adopted and reference is assumed to a theoretic framework that admits different degrees of membership to linguistic categories, many of the presumed abnormalities of evaluative morphology disappear.
4.1. Properties of Evaluative Suffixes
The main characteristics of evaluative affixes can be schematized as follows:
1. An evaluative affix can be attached to words belonging to different syntactic categories.
2. Evaluative affixes usually do not change the syntactic category of the base-word.
3. Evaluative affixes can change the subcategorization frame of the base
4. The application of evaluative affixes is usually constrained by restrictions.
5. There are two ways of interaction between different evaluative affixes:
i. On the syntagmatic layer: sequences of more evaluative affixes (or evaluative strategies) are widely attested;
ii On the paradigmatic layer: evaluative affixes violate the ‘Blocking Rule’ (cf. Scalise, Cerasa, Drigo, Gottardo, & Zannier, 1983); that is to say, more evaluative affixes sharing the same meaning (rival affixes) can be attached to same base-word.
6. Evaluative affixes are often preceded by interfixes, which can be explained only partially in terms of readjustment rules.
7. Evaluative affixes are not relevant for syntax:
i. They do not trigger any evaluative agreement;
ii They are not required by any syntactic context;
iii. They do not change the syntactic context of occurrence.
8. The meaning of evaluative affixes is usually not referential (that is, the base-word and the derived word usually have the same referent).
9. There is a hyponymy relation between a word with an evaluative affix and the base-word.
10. Evaluative affixes perform a modification function; their meaning can be foreseen according to four semantic values:
11. Evaluative complex words can undergo a lexicalization process
In the next paragraph, these properties are briefly surveyed. For further details, readers can refer to the first part of Grandi (2005) (which is the source of the data presented henceforth) and to Grandi (2015).
The properties in points 1 and 2 represent the so-called ‘categorial neutrality,’ undoubtedly the most problematic feature of the behavior of evaluative suffixes. In fact, they seem to select words belonging to different syntactic categories (especially nouns, adjectives, and verbs; but also other parts of speech such as numerals, determiners, etc.) without modifying the category of the base-word:
Such peculiar behavior in reality finds a convincing explanation on a semantic-functional perspective, as suggested by Bauer (1997, p. 549), “it does seem generally to be the case that evaluative morphology does not change the syntactic category of the base. This might be expected from the function of evaluative morphology. A noun which is noted as being of a particular size is still a noun; a noun which is stated to be liked or disliked is still a noun; an adjective which does not apply with its full force still remains an adjective.”
The hypothesis, advanced by several scholars, that the same neutrality can also be extended to the information contained in the sub-categorization frame is contradicted by the facts. Evaluative affixes often change the gender of the base (e.g., Bulgarian brat masc. ‘brother’ > bratlé ntr. ‘little brother’ or Portuguese abelha fem. ‘bee’ > abelhão masc. ‘large bee; hornet’). Moreover, diminutive affixes can attribute the feature [+ countable] to a mass noun (e.g., Berber xiz:ut ‘carrots’ > taxiz:utt ‘a carrot’). Furthermore, some evaluative affixes affect the actional information of the base.
The property in point 4 refutes another, diffused belief relative to evaluative affixes that, according to their application would be totally free, namely, not constrained by any restrictions. In reality, a recognition of the data is sufficient to realize, for example, how evaluatives are rarely formed from abstract and mass nouns. The central domain of evaluative affixes consists in fact of [+ concrete] and [+ countable] nouns, and therefore its definition is based on semantic restrictions. Also within their domain of application, the productivity of evaluative affixes is far from absolute. For example, Italian evaluative suffixes tend not to join to words when the final syllable sounds the same or similar to that of the suffix (e.g. *lettetto vs. lettino ‘small bed’ and *cuginino vs. cuginetto ‘little cousin’); they have a preference for simple words (e.g., Italian pizzetta ‘small pizza’ vs. ?pizzerietta ‘small pizzeria’ or *attaccapannino ‘small coat peg’); they usually exclude learned words or words that belong to elevated or formal registers (e.g., Italian mogliettina ‘little, dear wife,’ but *consortina ‘little, dear consort,’ pretino ‘small, young priest,’ but *reverendino ‘small, young reverend,’ etc.); they tend to exclude also words which go back to the so-called subordinate conceptual classes and to the super-ordinate conceptual classes (e.g., while gattino ‘small cat’ is an existing word in Italian, the diminutive of the superordinate term ?mammiferino ‘small mammal’ is rarely acceptable, as well as the diminutives of the subordinate terms such as ?siamesino, ?persianino o ?certosinetto, etc.). Similar restrictions are attested in other languages.
The properties indicated in point 5 can be exemplified by forms listed in (12):
Even if the reliability of the blocking rule has been denied (cf. for example, Bauer, Lieber, & Plag, 2013 for English data), it cannot be disclaimed that evaluative morphology’s inclination to interchange different strategies has no parallel in word formation.
The property 6 can sometimes be considered as the side effect of the property indicated in point 5.i in a sequence of evaluative affixes, the elements in the internal position tend to weaken in the phonetic body and in their semantic contribution, becoming interfixes (INTF in examples below). Although most interfixes derive from the transformation of evaluative affixes employed in sequence, also the use of interfixes for which it is not possible to reconstruct an origin of this type is widely documented:
For an accurate examination of the relationship between evaluative affixes and interfixes, see Dressler and Merlini Barbaresi (1989).
The properties in point 7 have, as we will see in the next paragraph, a great theoretical relevance. In fact, despite very few exceptions, evaluative affixes are never obligatory: evaluative affixes are not applied due to syntactic necessity, and no syntactic construction usually determines the use of an evaluative suffix, as the Italian sentences in (15) show:
In short, in the noun phrase, the use of the diminutive on the head does not trigger the additional use of the diminutive on the adjectival modifiers. Moreover, the diminutive can be applied to the modifiers, but not to the head. Therefore, evaluatives usually do not create evaluative agreement.
As far as semantics is concerned, in the previous section we have already said that the meaning of evaluative affixes is not referential, and we have already discussed the two dimensions of evaluative morphology (descriptive: SMALL vs. BIG; qualitative: GOOD vs. BAD). As far as the properties in point 9 are concerned, by saying that an evaluative form is a hyponym with respect to its own base, I mean that between a word and its evaluative form there is a relationship of implication: saying that X is a cagnolino ‘small dog’ implies that X is a cane ‘dog’ (while supposing that X is a canile ‘dog-kennel’ does not imply that X is a cane ‘dog’). This relationship could be formalized through the paraphrase ‘is a type of ’: a cagnolino is a type of cane (while a canile is not a type of cane).
4.2. Derivation of Inflexion?
After having presented the main properties of evaluative suffixes, it is necessary to briefly address the problem relating to their position within the morphological component, which, as said before, played a crucial role in the literature in the field.
A survey of the properties listed in the last paragraph does not give an encouraging result: some of them (the characteristics in points 3, 4, 7, and 11, for example) are typically derivational; others are inflexional (for example those in points 2, 8, and 9), and finally the properties indicated in 1, 2, 5, 6, but also 8, 9, and 10 cannot be attributed either to derivation or to inflexion. This situation seems to weaken the possibility of collocating evaluative affixes within the morphological component in a clear and unambiguous way. However, before discarding this possibility, it is necessary to assess whether the properties listed in section 4.1 form an internally homogeneous set; in other words, whether they all have the same ‘weight.’ I believe that the answer to this question must be negative: some of the properties of evaluative suffixes seem to play a more important role compared to the others. I am referring to the properties listed in point 7, relative to the relationship between evaluative affixes and syntax: evaluative affixes in most languages have no syntactic relevance, since they are never obligatory and do not trigger any kind of agreement. I believe that this characteristic plays a fundamental role in defining their collocation: evaluative affixes are in all ways derivational affixes. In fact, while inflexional affixes are obligatorily required by specific syntactic contexts and trigger off mechanisms of agreement, typically derivational affixes are always free and do not determine mutations in the syntactic context of occurrence.
This conclusion, however, must not encourage any neglect of the fact that few properties of evaluative affixes have an undeniable similarity with that of inflexional affixes. However, if inflexion and derivation come to be considered as the two extremes of a continuum and not, rather, as two sealed off compartments reciprocally independent (cf. Plank, 1994, p. 1672), the peculiar characteristics of evaluative affixes cease to be inconvenient abnormalities and become the most evident clues of their ‘peripheral’ position within derivation. In other words, as proposed by Dressler and Merlini Barbaresi (1992, p. 21), evaluative affixes, even though they are of derivational nature, cannot be considered as “a prototypical representative of derivational morphology.” That is equivalent to saying that evaluative suffixes are less derivational than other affixes (for example, affixes that create agent nouns), but not to the point of being defined as inflexional or being located in an autonomous and distinct class.
As mentioned above, this conclusion, though supported by an extremely vast cross-linguistic comparison, cannot be defined as a linguistic universal. In fact, there are languages in which evaluative suffixes have an eminently inflexional character. This is the case, for example, in Bantu languages, in which the addition of an evaluative prefix to the head of a noun phrase implies the addition of the same prefixes to all its modifiers too:
Bauer, L. (1997). Evaluative morphology: in search of universals. Studies in Language, 21(3), 533–575.Find this resource:
Dressler, W. U., & Merlini Barbaresi, L. (1994). Morphopragmatics: Diminutives and intensifiers in Italian, German and other languages. New York: Mouton de Gruyter.Find this resource:
Grandi, N. (2005). Sardinian evaluative morphology in typological perspective. In I. Putzu (Ed.), Sardinian in typological perspective (pp. 188–209). Bochum, Germany: Brockmeyer University Press. Discusses the place of evaluative morphology with respect to inflection and derivationFind this resource:
Grandi, N., & Körtvélyessy (Eds.), (2015). The Edinburgh handbook of evaluative morphology. Edinburgh: Edinburgh University Press. This is the best starting point for readers who want to read more deeply on the topic discussed in this article. The first part of the handbook deals with some of the theoretical issues discussed in the article, including the definition of evaluative morphology. In the second part of the volume, readers will find more than 50 monographic descriptions of single languages.Find this resource:
Jurafsky, D. (1993). Universals in the semantics of the diminutive. Proceedings of the Nineteenth Annual Meeting of the Berkeley Linguistics Society, 19(1), 423–436.Find this resource:
Jurafsky, D. (1996). Universal tendencies in the semantics of the diminutive. Language, 72(3), 533–578.Find this resource:
Körtvélyessy, L. (2015). Evaluative morphology from a cross-linguistic perspective. Newcastle upon Tyne, U.K.: Cambridge Scholars Publishing. A general survey of this research field.Find this resource:
Merlini Barbaresi, L. (2015). Evaluative morphology and pragmatics. In N. Grandi, & L. Körtvélyessy (Eds.), The Edinburgh handbook of evaluative morphology (pp. 32–42). Edinburgh: Edinburgh University Press. Addresses the pragmatics of evaluative morphology.Find this resource:
Mutz, K. (2015). Evaluative morphology in a diachronic perspective. In N. Grandi, L. Körtvélyessy (Eds.), The Edinburgh handbook of evaluative morphology (pp. 142–154). Edinburgh: Edinburgh University Press. Evaluative morphology from the point of view of language change.Find this resource:
Prieto, V. M. (2015). The semantics of evaluative morphology. In N. Grandi, L. Körtvélyessy (Eds.), The Edinburgh handbook of evaluative morphology (pp. 21–31). Edinburgh: Edinburgh University Press. Expounds on the semantics of evaluative morphology.Find this resource:
Scalise, S. (1984). Generative morphology. Dordecht, The Netherlands: Foris. All scholars interested in this field cannot forget this pioneering work.Find this resource:
Schneider, K. P. (2013). The truth about diminutives, and how we can find it: Some theoretical and methodological considerations. SKASE Journal of Theoretical Linguistics, 10(1). A recent work on evaluative morphology, with a special focus on the pragmatic dimension.Find this resource:
Stump, G. T. (1993). How peculiar is evaluative morphology? Journal of Linguistics, 29, 1–36.Find this resource:
Bauer, L. (1997). Evaluative morphology: In search of universals. Studies in language, 21(3), 533–575.Find this resource:
Bauer, L., Lieber, R., & Plag, I. (2013). The Oxford reference guide to English morphology. Oxford: Oxford University Press.Find this resource:
Beard, R. (1981). The Indo-European lexicon: A full syncronic theory. Amsterdam: North Holland.Find this resource:
Carstairs-McCarthy A. (1992). Current morphology. New York: Routledge.Find this resource:
Dressler, W. U., & Merlini Barbaresi, L. (1989). Interfissi e non-interfissi autosuffissali nell’italiano, spagnolo e inglese. In F. Foresti, E. Rizzi, & P. Benedini (Eds.), L’italiano tra le lingue romanze (pp. 243–252). Atti del XX Congresso della Società di Linguistica Italiana. Rome: Bulzoni.Find this resource:
Dressler, W. U., & Merlini Barbaresi, L. (1992). Italian diminutives as non-prototypical word formation. In L. Tonelli, & W. U. Dressler (Eds.), Natural morphology: Perspectives for the nineties (pp. 21–30). Padua, Italy: Unipress.Find this resource:
Dressler, W. U., & Merlini Barbaresi, L. (1994). Morphopragmatics: Diminutives and intensifiers in Italian, German, and other languages. New York: Mouton de Gruyter.Find this resource:
Grandi, N. (2002). Morfologie in contatto. Le costruzioni valutative nelle lingue del Mediterraneo. Milan, Ialy, Franco Angeli.Find this resource:
Grandi, N. (2005). Sardinian evaluative morphology in typological perspective. In I. Putzu (Ed.), Sardinian in typological perspective (pp. 188–209). Bochum, Germany: Brockmeyer University Press.Find this resource:
Grandi, N. (2015). The place of evaluation within morphology. In Edinburgh handbook of evaluative morphology (pp. 74–90). Edinburgh: Edinburgh University Press.Find this resource:
Grandi, N., & Körtvélyessy, (2015a). Introduction: Why evaluative morphology? In N. Grandi, & L. Körtvélyessy (Eds.), The Edinburgh handbook of evaluative morphology (pp. 3–20), Edinburgh: Edinburgh University Press.Find this resource:
Grandi, N., & Körtvélyessy, L. (Eds.), (2015b). The Edinburgh handbook of evaluative morphology. Edinburgh: Edinburgh University Press.Find this resource:
Haspelmath, M. (2003). The geometry of grammatical meaning: Semantic maps and cross-linguistic comparison. In M. Tomasello (Ed.), The new psychology of language (Vol. 2, pp. 211–242). Mahwah, NJ: Erlbaum.Find this resource:
Haspelmath, M. (2007). Pre-established categories don't exist: Consequences for language description and typology. Linguistic Typology, 11(1), 119–132.Find this resource:
Haspelmath, M. (2010). Comparative concepts and descriptive categories in cross-linguistic studies. Language, 86(3), 663–687.Find this resource:
Jurafsky, D. (1993). Universals in the semantics of the diminutive. Proceedings of the Nineteenth Annual Meeting of the Berkeley Linguistics Society, 19(1), 423–436.Find this resource:
Jurafsky, D. (1996). Universal tendencies in the semantics of the diminutive. Language, 72(3), 533–578.Find this resource:
Körtvélyessy, L. (2015). Evaluative morphology from a cross-linguistic perspective. Newcastle upon Tyne, U.K.: Cambridge Scholars Publishing.Find this resource:
Merlini Barbaresi, L. (2015). Evaluative morphology and pragmatics. In N. Grandi, L. Körtvélyessy (Eds.), The Edinburgh handbook of evaluative morphology (pp. 32–42). Edinburgh: Edinburgh University Press.Find this resource:
Mutz, K. (2015). Evaluative morphology in a diachronic perspective. In N. Grandi, L. Körtvélyessy (Eds.), The Edinburgh handbook of evaluative morphology (pp. 142–154), Edinburgh: Edinburgh University Press.Find this resource:
Plank, F. (1994). Inflection and derivation. In R. E. Asher, J. M. Y. Simpson (Eds.), Encyclopedia of language and linguistics (Vol. 3, pp. 1671–1678). Oxford: Pergamon Press.Find this resource:
Prieto, V. M. (2005). Spanish evaluative morphology: Pragmatic, sociolinguistic, and semantic issues. Doctoral diss. University of Florida.Find this resource:
Prieto, V. M. (2015). The semantics of evaluative morphology In N. Grandi, L. Körtvélyessy (Eds.), The Edinburgh handbook of evaluative morphology (pp. 21–31). Edinburgh: Edinburgh University Press.Find this resource:
Scalise, S. (1984). Generative morphology. Dordecht, The Netherlands: Foris.Find this resource:
Scalise, S. (1994). Morfologia. Bologna, Italy: il Mulino.Find this resource:
Scalise, S., Cerasa, M., Drigo, M., Gottardo, M., & Zannier, I. (1983). Sulla nozione di Blocking in morfologia derivazionale. Lingua e Stile, 2, 243–269.Find this resource:
Schneider, K. P. (2013). The truth about diminutives, and how we can find it: Some theoretical and methodological considerations. SKASE Journal of Theoretical Linguistics, 10(1).Find this resource:
Simon, H. J., & Wiese, H. (2011). What are exceptions? And what can be done about them? In H. J. Simon & H. Wiese (Eds.), Expecting the unexpected: Exceptions in grammar (pp. 3–31). Berlin: Mouton De Gruyter.Find this resource:
Štekauer, P. (2015). Word-formation processes in evaluative morphology. In N. Grandi & L. Körtvélyessy (Eds.), The Edinburgh handbook of evaluative morphology (pp. 43–60). Edinburgh: Edinburgh University Press.Find this resource:
Stump, G. T. (1993). How peculiar is evaluative morphology? Journal of Linguistics, 29, 1–36.Find this resource:
Szymanek, B. (1988). Categories and categorization in morphology. Lublin, Poland: Redakcja Wydawnictw Katolickiego Uniwersytetu Lubelskiego.Find this resource:
Wierzbicka, A. (1996). Semantics: Primes and universals. Oxford: Oxford University Press.Find this resource: