Skip to navigation – Site map
Papers

Renewal and Innovation in the Emergence of Indo-European Evaluative Morphology

Nicola Grandi

Abstract

A diachronic survey in the field of the so-called evaluative morphology in some branches of the Indoeuropean family (above all Romance and Slavonic languages and Greek) reveals two different tendencies. On the one side suffixes that displayed a diminutive value in the earliest stages of these languages do not correspond to present-day diminutive suffixes. On the other side, Proto-Indoeuropean before and Latin and Ancient Greek then lacked augmentative suffixes at all, while Romance languages and Modern Greek have at their disposal some of them. So, diminutives seem a dynamic and unstable linguistic strategy, which, in the course of ages, has undergone a wide (cyclic?) renewal: the semantic function has been kept on, while the formal strategies to express it have changed. Instead, augmentatives seem to be the result of an innovation: to a sure point, a new category has been introduced and each language has had to find the means to express it. In a diachronic perspective, augmentatives seem to be a more steady linguistic strategy than diminutives. In this paper I intend at reconstructing, going backwards, the genesis of some Romance, Slavonic, and Greek diminutive and augmentative suffixes in order to single out both their semantic archetypes and possible common stages recurring in their evolutive processes.

Top of page

Full text

I thank Emanuele Banfi, Sandro Caruana and Fabio Montermini, who read and commented a previous version of this paper.

1. Theoretical background

  • 1 In this paper, by the label ‘evaluative morphology’ I will refer only to its descriptive/quantitati (...)

1A diachronic survey in the field of the so-called evaluative morphology1 of modern Indo-European languages spoken in Europe reveals two different tendencies. On the one side suffixes that displayed a diminutive value in the earliest stages of these languages usually do not correspond to present-day diminutive suffixes. For example, actual Romance diminutive suffixes did not have a diminutive meaning in Latin, while Latin diminutive suffixes have lost this value or are not used any longer in Romance languages. On the other side, both Proto-Indo-European and ancient Indo-European languages (attested or reconstructed: Latin, Ancient Greek, Proto-Slavonic, Proto-Germanic, Proto-Celtic, etc.) did not have augmentative suffixes at all, while many modern languages have some of them at their disposal (this is the case, for example, of most Romance and Slavonic languages and of Modern Greek).

2So, in the history of Indo-European evaluative morphology there seems to have been an interaction between a process of renewal, concerning diminutives, and a process of innovation, concerning augmentatives.

3In this paper I intend to reconstruct, by investigating the issue from a historical point of view, the genesis of some Indo-European diminutive and augmentative suffixes, in order to single out both their semantic archetypes and possible common stages recurring in their development. My focus will be mainly on Romance languages, Slavonic languages and Greek, but data from Celtic, Germanic and Baltic languages will also be presented and discussed. Some typologically different and non-Indo-European languages, such as Berber, Vietnamese, Malay, Thai and Hmong, will be referred to in order to provide terms of comparison.

4The theoretical framework that, in my opinion, provides the best tools for a satisfactory explanation of the intricate history of evaluative morphology is the well-known ‘dynamic typology’. As suggested by Greenberg [1969, 1978, 1995], the linguistic change cannot just be seen as a mere sequence of origins and losses of single linguistic items or functions (where “origins seem inherently more interesting” Greenberg [1995: 149]), but as the symptom of a slow and gradual shift from a typological state to another one; in such a view, origins and losses are equally important.

5In my opinion, the vicissitudes of Indo-European evaluative morphology represent a promising field for the assumptions of this theoretical framework.

6In the following sections of this paper, after a few introductory remarks (§ 2), we will carry out an in-depth diachronic investigation of augmentative (§ 3) and diminutive (§ 4) suffixes of the previously mentioned Indo-European languages. Then, these data will be compared to data from non-Indo-European languages (§ 5), in order to establish whether universal or at least widely shared tendencies in the development of evaluative morphology do exist or not (§ 6).

2. Stability and frequency in evaluative morphology

7If, as we stated in § 1, the history of evaluative morphology is to be explained within a typological approach, we must select the types that occur among the languages included in my sample. By means of the combination of the two parameters chosen for this analysis, that is ‘presence/absence of diminutives’ and ‘presence/absence of augmentatives’, four possible types can be singled out:

(1)

Type A: presence of diminutives; absence of augmentatives;

Type B: presence of both diminutives and augmentatives;

Type C: absence of both diminutives and augmentatives;

Type D: absence of diminutives; presence of augmentatives;

  • 2 To my knowledge, no counterexample to this state of affairs has been found yet.

8However, it has been noted that “augmentatives represent a marked category opposed to the unmarked category of diminutives” (Dressler / Merlini Barbaresi [1994: 430]). The markedness of augmentatives is confirmed by their being cross-linguistically less common than diminutives. This situation has been represented by an implicational correlation, which is supposed to be universal:2 augmentatives ⊃ diminutives (cf., among others, Haas [1972]). This correlation must be read as follows: if a language has some morphological devices to form augmentatives, then it must have morphological diminutives too, but not vice versa. By means of this implication, the type D in (1) is ruled out. So, as to descriptive/quantitative evaluative morphology, only types A, B and C can actually occur. A survey of the morphological inventories of modern Indo-European languages of Europe supports this assumption:

  • 3 The notation A (> B?) indicates that in French an augmentative suffix (-ard; cf. meulard ‘big mills (...)
  • 4 The question mark indicates that English displays some diminutive suffixes (for example -let: pigle (...)
  • 5 As in the case of French, the notation A ( B?) indicates that in Slovene a few augmentative suffixe (...)

(2)

Romance languages

Portuguese

type B

Spanish

type B

Catalan

type B

Occitan

type A

French

type A (> B?)3

Italian

type B

Sardinian

type A

Rumanian

type B

Germanic languages

English

type C (?)4

German

type A

Dutch

type A

Swedish

type C

Danish

type C

Celtic languages

Breton

type A

Irish

type A

Scottish Gaelic

type A

Welsh

type A

Albanian

Albanian

type B

Greek

Greek

type B

Slavonic languages

Slovene

type A (> B?)5

Serbo-Croat

type B

Bulgarian

type B

Macedonian

type B

Russian

type B

Polish

type B

Czech

type B

Baltic languages

Latvian

type B

Lithuanian

type B (cf. Grandi [2003])

9So, modern Indo-European languages of Europe are equally distributed between type A and type B, the only exceptions being Scandinavian languages and possibly English. In this sense, Europe reproduces on a smaller scale a worldwide situation: a quick glance at diminutives and augmentatives in languages other than Indo-European shows that types A and B are largely prevalent. For example, the great majority of Afroasiatic languages (except for Berber, Maltese and Moroccan Arabic) are of type A; almost all Bantu languages can be assigned to type B; most Sino-Tibetan languages are of type C.

10If we consider diachronic aspects, the picture changes roughly, since in ancient Europe type B was not attested at all. Latin, Ancient Greek, Common Slavonic, Common Germanic and Common Celtic can all be assigned to type A and in this respect they are consistent with Proto-Indo-European. In fact, Proto-Indo-European displayed two morphemes, the well-known *-lo- and *-ko-, with a possible diminutive meaning, but had no morphological augmentative.

11Within the theoretical framework of the so-called ‘dynamic typology’ the distribution of linguistic types is determined by two independent, but complementary factors: stability and frequency. The former indicates “the probability that a language which is in a particular state will exit this state”, while the latter indicates “the probability that a language will enter a particular state” (Greenberg [1995: 151]). In other words, a high degree of stability corresponds to a probable likelihood for a state to be preserved by a language or by a group of languages; a high degree of frequency corresponds to a probable likelihood for a state to be assumed by a language or by a group of languages.

12Stability and frequency determine the areal and genetic spread of linguistic states: a stable state will show a uniform diffusion within genetic groups, since it is usually inherited from a common ancestor, whereas a frequent state will tend to be areally widespread, but sporadic within linguistic families.

  • 6 According to Greenberg [1995: 152] the combination of the highest stability and the highest frequen (...)

13If we read the list in (2) in the light of these two parameters, we may note that diminutives exhibit a high value both for stability and frequency: they are areally widespread and common to related languages.6 However, in this case we can assert that wide areal diffusion is reasonably a consequence of the fact that they have been inherited from a common ancestor (for example, we can state that diminutive suffixes are attested in all Romance languages because the semantic category ‘diminutive’ has been inherited from Latin; the same holds for the Slavonic group). So, as far as diminutives are concerned, it is stability that plays a crucial role. On the contrary, augmentatives are unstable, but quite frequent: their occurrences are regular in clear cut areas (see, for example, the Ibero-Romance zone, the Balkans, etc.), but not among related languages (for example, only some Romance languages have augmentative suffixes: the semantic category ‘augmentativehas not been inherited from Latin). So, in this case frequency clearly prevails over stability.

14In my opinion, the use of the two parameters of stability and frequency has relevant consequences also for the investigation of the history of linguistic states, allowing us to make strong predictions about the nature of linguistic changes by means of which they have been produced. My hypothesis is that it is plausible for a state which is widespread both areally and genetically and which is shared by the great majority of the members of many linguistic families (in other words, for a stable and frequent state) to be the consequence of a development which always proceeds along the same course (i.e. the same starting point, the same intermediate stages, the same result), independently of the genetic relationships among the languages involved and of their geographical position.

15On the other hand, it is plausible for a state which is attested in different and not contiguous areas, in which unrelated languages are spoken (that is, for a frequent, but unstable state) to be the consequence of many areal-specific processes.

16To sum up, my hypothesis is that the genesis of linguistic states in which stability is prevalent tends to be conditioned by very general typological tendencies, while the development of states in which frequency is prevalent is significantly constrained by areal factors.

17If this is true, by referring to evaluative morphology, we can hypothesize that the development (and the subsequent renewal) of diminutive suffixes has taken place according to a unique general typological tendency in all their occurrences, while augmentative suffixes have emerged following different and areal-specific evolutive paths. The data we will present in the next sections seem to support this hypothesis.

3. Augmentative suffixes in Indo-European languages of Europe

18The list in (2) shows that among the modern Indo-European languages of Europe most Romance languages, many Slavonic languages, Baltic languages and Greek have undergone the process of innovation which has led to the emergence of augmentative suffixes. In Romance and Slavonic languages and in Greek four different evolutive paths are attested. Two of them occur both in the great majority of Romance languages and in Greek (cf. §§ 3.1.1 and 3.1.2); one involves just a few Romance languages (§ 3.1.3), and one seems to be peculiar to the Slavonic region (§ 3.2).

3.1. Romance languages and Greek

19As we indicated before, in Romance languages (with the exception of Gallo-Romance languages and Sardinian) and in Greek, three different semantic shifts seem to have triggered the development of augmentative suffixes. Two of them are common to most Romance languages and Greek; a third one took place only in Portuguese, in Spanish and in Southern Italian.

3.1.1. Agentive / pejorative > augmentative

  • 7 In this section we summarize the picture already presented in Grandi [2002b], to which we refer the (...)

20In this section I will draw a parallel sketch of the evolution of the most widespread Romance augmentative suffixes (It. -one in gattone ‘big cat’, Sp. -ón in hombrón ‘hulk of a man’, Port. -ão in ceirão ‘large woven basket’), which are the outcome of a single Latin suffix, ‑(i)o, ‑(i)ōnis, and of the Modern Greek augmentative suffix -άς (κεφαλάς ‘big head’), which is the result of two closely related Ancient Greek suffixes, ‑ᾶς and ‑ίας.7

21As we saw in § 2, neither in Latin nor in Ancient Greek ‘typical’ augmentative suffixes were attested. However, both Latin and Ancient Greek had some very productive derivational suffixes at their disposal, Lat. ‑(i)o, ‑(i)ōnis and A.Gr. ‑ς and ‑ίας, attested in different formations with different semantic readings. In the great majority of their occurrences, these suffixes were used to form masculine animate nouns designating human beings with a particular, often physical, characteristic or with the habit of performing an action in an exaggerated way:

  • 8 Latin data are from Lazzeroni [1963] and Gaide [1988]; Ancient Greek data are from Chantraine [1933 (...)

(3)

a. N[‑anim]-PEJ > N[+ hum].M8

Lat.

gānĕ(a)-o
tavern:N[- anim].F-PEJ.N[+ hum].M / ’guzzler, a dissolute person’

ment(um)-o
chin:N[- anim].NTR-PEJ.N[+ hum].M / ’person with protruding chin’

A.Gr.

κορυζ(α)-ᾶς
snot, mucus from the nose:N
[- anim].F-PEJ.N[+ hum].M / ’whipper-snapper’

πωγων-ίας
beard:N
[- anim].M-PEJ.N[+ hum].M / ’bearded person’

 

 

b. V-PEJ > N[+ hum].M

Lat.

err(āre)-o
to ramble, to wander about:V-PEJ.N
[+ hum].M / ’rambler, vagabond, wanderer; deserter’

mand(ĕre)-o
to bite, to chew:V-PEJ.N
[+ hum].M / ’hearty eater’

A.Gr.

καταφαγ(εῖν)-ᾶς
to eat, to devour, to squander:V-PEJ.N[+ hum].M / ’squanderer’

 

 

c. A-PEJ > N[+ hum].M

Lat.

mīr(us)-ĭo
wonderful, marvellous:A-PEJ.N
[+ hum].M / ’monster’

miscell(us)-io
mixed, various:A-PEJ.N
[+ hum].M / ’muddler, bungler’

A.Gr.

γυναικ(εῖος)-ίας
womanly, feminine:A-PEJ.N[+ hum].M / ’womanish man’

22The semantic reading of data in (3) can be brought back to the paraphrase ‘one who is/makes/has X to a high degree’. The use of these suffixes in onomastics, and specifically in the formation of cognomina, proper names of mask-characters and nicknames, is a consequence of their pejorative and caricatural meaning:

  • 9 As usual, a particular characteristic gives rise to a nickname that afterwards can ‘crystallize’; i (...)

(4)

Latin:

a. cognomina:

Nās(us)-o
nose:N[- anim].M-PEJ.N[+ hum].M / ’Nasone’ (lit. ‘who has a particular nose’)

Pēs
‘foot’ > Pĕd-o
foot:N[- anim].M-PEJ.N[+ hum].M / ’Pedone’ (lit. ‘flat-footed person’)

 

b. proper names of mask-characters and actors:

Bucc(a)-o
mouth; cheek:N[- anim].F-PEJ.N[+ hum].M / ’Bucco’ (a character of fabulae Atellanae; lit. ‘silly, insolent’)

Turp(is)-ĭo
ugly, repulsive:A-PEJ.N[+ hum].M / ’Turpio’ (name of a comic actor)9

 

(5)

Ancient Greek:

a. nicknames:

Μηνόδωρος > Μηνᾶς

Νικομήδης > Νικομᾶς

 

b. epithets:

Λοξ(ός)-ίας
oblique, ambiguous:A-PEJ.N[+ anim].M / ’oblique’ (epithet of Apollon; probably because of the ambiguity of his oracles)

 

c. Proper Names:

Ἇμυν(α)-ίας

defence:N[- anim].F-PEJ.N[+ hum].M / ’Aminia’ (lit. ‘ready to defend himself’)

  • 10 For Latin, cf. Gaide [1988: 126]: “[l]e sème augmentatif (« virtuel ») que présente le suffixe dans (...)

23The transfer to the evaluative function is intuitively clear: the suffix no longer designates the possessor of an unusual property, but it identifies the property itself.10 So, a word such as Latin căpĭto (from the noun caput ‘head’) originally indicated a ‘big-headed person’ and, then, in Late Latin, just a ‘big head’. So, Romance and Modern Greek augmentative suffixes are the result of derivational suffixes originally used to form animate (often human) nouns with an agentive / pejorative / caricatural meaning; this original meaning is still preserved in modern languages, besides the ‘new’ augmentative meaning:

(6)

a. original pejorative/caricatural meaning

b. new augmentative meaning

It.

fif(a)-one
fear:N[- anim].F-AUG.N[+ hum].M
‘a cowardly person’

barc(a)-one
boat:N[- anim].F-AUG.N[- anim].M
‘big boat’

Sp.

com(er)-il-on
to eat:V-INTF-AUG.N[+ hum].M
‘hearty eater’

caj(a)-ón
case:N[- anim].F-AUG.N[- anim].M
‘large case’

Port.

beat(o)-ão
blessed:A-AUG.N
[+ hum].M
‘great hypocrite’

aban(o)-ão
shock:N
[-anim].M-AUG.N[-anim].M
‘great shock’

M.Gr.

καθίστ(ος)-ακας
seated:A-AUG.N[+ hum].M
‘lazy-bones’

κλεφτ(ης)-αράς
thief:N[+ hum].M-AUG.N[+ hum].M
‘big thief’

3.1.2. Collective > augmentative

24In Modern Greek and in some Romance languages (the languages spoken in the Western side of Roman zone, assuming the traditional and conventional division based on the Spezia-Rimini line) the augmentative meaning can be expressed simply by changing a masculine or neuter noun into feminine:

  • 11 Data of Romance languages are from Kahane / Kahane-Toole [1948-1949] and Volpati [1955].
  • 12 As for Italian, cf. also the very widespread couple pennello masc. ‘brush’ / pennellessa fem. ‘flat (...)
  • 13 Dialectal varieties.

(7) 11

a. M.Gr.

κεφάλι
head:NTR

>

κεφάλα
head:F / ’big head’

φίδι
snake: NTR

>

φιδάρα
snake:F / ’big snake’

b. It.

buco
hole:M

>

buca
hole:F / ’cave, hollow pit’

fiasco
flask:M

>

fiasca
flask:F / ’large flask’
12

c. Fr.13

pré
meadow:M

>

prée (Western French)
meadow:F / ’wide meadow’

sac
bag:M

>

sache (Central French)
bag:F / ’large-sized bag’

d. Cat.

pas
step:M

>

passa
step:F / ’long step

plat
plate:M

>

plata
plate:M / large plate’

e. Sp.

cesto
basket:M

>

cesta
basket:F / ’large hand basket for dirty linen’

garbanzo
chick-pea:M

>

garbanza
chick-pea:F / ’large type of chick pea’

f. Port.

rato
mouse:M

>

rata
mouse:F / ’rat’

caldeiro
cauldron:M

>

caldeira
cauldron:F / ’large boiler’

25As to Romance forms, Kahane and Kahane-Toole [1948-1949: 154-155] state that

as a rule, the quantitative -o ~ -a contrast applies to strikingly three-dimensional objects. […]. Since the -o ~ -a relation as an expression of sex contrast refers particularly to human beings, it is never used to express difference in size between human beings. It is, in general, limited to the inanimate. It may express difference in size between animals only where difference in sex is unrecognizable or irrelevant or where the female is larger than the male.

26If we focus on Modern Greek data, the origin of feminine augmentatives becomes evident. The Greek augmentative suffix -α is characterized by three specific properties:

i] it forms only feminine nouns;
ii] it cannot be used to form animate nouns from inanimate nouns, adjectives or verbs;
iii] it is frequently (but not exclusively) attached to neuter nouns.

  • 14 See Heilmann [1963: 149]: “La desinenza -α del nom. voc. acc. neutro (ζυγά, μέτρα) rappresenta lo s (...)
  • 15 We must be very cautious about projecting this synchonic generalization on a diachronic dimension. (...)

27These properties, as well as the phonological form of the suffix, can be traced back to Ancient Greek and Latin neuter plural endings. As it is well known, the original meaning of these endings was collective.14 It is plausible for collective to have been the semantic archetype of present-day feminine augmentatives, since these two meanings are undoubtedly very close. Moreover, a parallel survey of evaluative morphology and category of number reveals a wider and more systematic interaction between them. The synchronic link between diminutives and singulatives has been stressed quite often: there exists a strong cross-linguistic tendency to express these two meanings by means of the same formal item.15 In this picture, a relation between augmentatives and collectives would certainly be feasible (Cf. Grandi [2001]).

3.1.3. Relational > augmentative

28In Latin the relational suffix ‑ācĕus (a, um) was productively used to form adjectives indicating similarity, material source or provenience, and approximation:

(8)

charta
‘paper’

>

chartācĕus (a, um)
‘made of paper’

folium
‘leaf’

>

foliācĕus (a, um)
‘leaf-shaped’

29Sometimes this semantic value is still preserved among Romance languages, but if this is the case, the suffix does not undergo the expected phonological changes and exhibits a ‘learned shape’ (that is, a similarity with the Latin suffix: It. cartaceo ‘made of paper’; the regular Italian form of the suffix might be -accio).

30According to Rohlfs [1969: 366], the idea of similarity gave rise to the pejorative meaning which is largely prevalent in Italian words ending in -accio (tavolaccio ‘plank-bed’, ragazzaccio ‘naughty boy’, figuraccia ‘(to make) a bad impression’). Rholfs states also that the suffix “può esprimere – oggettivamente – qualcosa di rozzo o di più grande, oppure – soggettivamente – ciò che è meno buono.” In this way, he explains augmentative forms such as Port. animalaço ‘big animal’ from animal and barbaça ‘long beard’ from barba or Cat. calorassa ‘strong heat’ from calor ‘heat’. A similar semantic reading occurs in some dialects of Southern Italy: canazzu (‘big dog’ – Calabria), vuccazza (‘big mouth’ – Calabria), festazza (‘great party’ – Naples), doddazza (‘big dowry’ – Abruzzo).

31Augmentative outcomes of Latin -ācĕus (a, um) are quite rare in the Romance area as they are attested only in Ibero-Romance zone and in the Southern part of Italy.

3.2. Slavonic languages: locative > augmentative

32Unlike Romance languages and Greek, which make use of few augmentative suffixes, Slavonic languages display rich and complex inventories of augmentative suffixes. For the topic being discussed in this paper, the most interesting are the related suffixes -ište/ -išče/-isko, the occurrences of which can be extensively traced along all the documentary tradition of Slavonic languages.

33According to Vaillant [1974: 422], these suffixes are the outcomes of a single Common Slavonic locative suffix: “en vieux slave ce suffixe se tire en principe de substantifs et fournit des noms qui indiquent le lieu.” Locative meaning was widely attested in Old Church Slavonic (cf. 9) and has been retained by modern Slavonic languages (cf. 10):

  • 16 Nandriş [19652: 88].

(9)16

OCS

sód(ŭ)-ište
judgment-LOC / ’court of law’

grob(ŭ)-ište
grave-LOC / ’cemetery’

 

(10)

Rus.

gúl’bišče
‘(public) walk’

Slov.

Brodíšče
‘place from which ferry-boats pass’

34But in modern Slavonic languages such as Bulgarian, Macedonian, Serbo-Croat, Russian, Polish and Czech the main and most frequent semantic value of the suffixes ‑ište/-išče/-isko is augmentative:

(11)

Blg.

det(é)-ište
child-AUG / ’big child’

žen(á)-íšte
woman-AUG / ’big, hefty woman’

Mac.

Čovek
‘man’

>

čoveč-ište

man-AUG / ’huge man’

  • 17 We can put forward the hypothesis that a possible link between locative and augmentative meaning is (...)
  • 18 On the contrary, as in the case of the semantic change presented in § 3.1.1, a form such as Late La (...)

35It is not easy to understand how the augmentative meaning might have developed from the original locative interpretation. In fact, it is necessary to point out that, unlike the semantic shifts we have seen in §§ 3.1.1, 3.1.2, and 3.1.3, a link between locative and augmentative meaning can hardly be found.17 Furthermore, in available records, the ‘new’ augmentative meaning seems to have emerged rather abruptly, as it had not been preceded by forms in which both semantic readings are possible.18

36Therefore, since it is impossible to single out the intermediate stages of this evolutive path, we have to limit ourselves to focusing on its areal-specific character: in fact a shift from a locative meaning to an augmentative does not seem to have occurred in other Indo-European languages.

37The data presented so far seem to support the hypothesis presented in § 2: in the history of a typological state in which frequency clearly prevails over stability areal constraints play a crucial role. The main augmentative suffixes attested in the European languages of Indo-European family are the result of different areal-specific evolutive paths, as can be seen in the following map:

4. Diminutive suffixes in Indo-European languages of Europe

38According to the list drawn in (2), diminutive suffixes display a high degree of stability in the terms provided by Greenberg: they are uniformly widespread in all genetic groups we have investigated, with the partial exception of languages belonging to the northern branch of the Germanic family.

39The issue of the origin of diminutives has been widely discussed by scholars. Today, the great majority of them agree in asserting that animate nouns have played a crucial role in the development of diminutive suffixes. Nevertheless, the origin of the real semantic archetype is still unclear and so is the procedure by means of which the diminutive value developed from it. Two assumptions seem prevalent in the literature: on the one hand, some scholars place the hypocoristic value as the starting point of the entire process; on the other hand, other scholars consider the designation of the genealogical relation between father and child (in the case of human beings) and/or between the adult and the young (in the case of animals) to be the semantic archetype of the diminutive value. Data from Romance languages (§ 4.1), from Greek (§ 4.2), and from Slavonic languages (§ 4.3) give evidence in favor of this second hypothesis: the shift in meaning of It. -ino (in tavolino ‘small table’), Gr. -άκι (in λαθάκι ‘a pardonable error’), Blg. -ec/-íca (in brátec ‘little brother’ / zeníca ‘little woman’), etc. towards the current diminutive value took its first steps in the designation of the genealogical relation between father and son and between the adult and the young. If we assume the meaning ‘child/young of…’ to be the semantic archetype, the emergence of the diminutive value seems easy to explain. In this picture, the expressive or hypocoristic nuances are not to be considered as original, but as a secondary consequence of this semantic change. As we will show, a swift glance both at other Indo-European groups (Germanic and Baltic), and to non-Indo-European languages (see § 5) seems to support this hypothesis.

4.1. Romance languages

  • 19 For an exhaustive survey of Latin outcomes of Indo-European denominal suffixes *‑ĭno-, *-īno- and * (...)

40Some of the most widely used Romance diminutive suffixes, It. -ino, sp. –ín, port. ‑inho, are the outcome of a single Latin suffix, -īnus, the various semantic readings of which are usually traced back to a vague relational value (cănīnus ‘pertaining to the dog’; vespertīnus ‘happening in the evening’, Sābātīnus ‘of Sabate, a town in Etruria’, etc.).19

41According to Butler [1971: 22-23], clear traces of the semantic reading ‘young X, child X’ can be discovered among the very first occurrences of the suffix, above all in its use in anthroponomy:

  • 20 Data and translations from Oxford Latin Dictionary (cf. Glare [1982]).

(13)

Agrippīna
‘the daughter of M. Vipsanius Agrippa’

Messālīnus
‘borne by M. Valerius Messalla Messalinus, son of M. Valerius Messalla Corvinus’20

42Another very interesting form in this respect is amitīnus ‘cousin’, whose literal meaning is ‘son of paternal aunt’, from amita (‘paternal aunt, father’s sister’).

  • 21 Cf. also Rohlfs [1969: 412]: “ciò che è meno compiuto è più piccolo”.

43Hakamies [1951: 9] states that “l’adulte est le prototype d’une espèce; par conséquent ce qui ou celui qui ressemble à l’espèce sans atteindre toutefois au prototype ne peut être que plus petit”; in this way he identifies the link between the original relational meaning and the diminutive interpretation.21 The derivational history of a form such as Lat. castŏrīnus / It. castorino can be represented as follows:

14)

Latin

Italian

castor]N
‘beaver’

>

castŏrīnus]A
‘resembling a beaver’

>

castŏrīnus]N>
‘young beaver’

castorino]N
‘small/little beaver’

44In the form at the end of this sequence, any reference to age has been lost.

45On the formal ground, this semantic shift comes about by means of a process of conversion (noun > adjective) placed between the second and the third stage of the sequence in (14).

46As always happens, a linguistic change spreads slowly and step by step. Therefore, it is not surprising to find occurrences of both the two semantic functions involved in the change in the same synchronic stage. So, for example, in Italian spoken approximately in the XII-XIII centuries, the suffix -ino was still attested both in anthroponomy (15) and in diminutive forms of animate nouns (16):

  • 22 F. = son.

(15)

i.

‘Giovannino f. Giovani’22 [sic] and ‘Lanfra(n)chino […]di s(er) Lanfra(n)co’
Memoriale dei camarlinghi del Ceppo dei poveri di Prato (1296-1305)

ii.

‘Ciampolino Ciampoli’
A document from Siena written between 1294 and 1375

iii.

‘Masino di Maso’
Quaderno dei creditori di Taddeo dell’Antella e compagni
(1345)

iv.

‘Mannellino de’Mannelli’
in D.Velluti, La cronica domestica (1367-70)

 

(16)

i.

‘Onde vedemo li parvuli desiderare massimamente un pomo; e poi, più procedendo, desiderare uno augellino.’
Dante, Convivio (IV, cap. 12), 1304-7

ii.

‘A la perfine, vinto per prieghi, acconsentìo al disiderio del re e battezzando il fanciullo, tacendo tutti quelli ch’erano presenti, il fanciullino rispuose: “Amen”.’
Leggenda Aurea
(XIV sec)

iii.

‘e, per più loro sicurtade, Medea ne menò seco uno suo fratellino: onde ella, essendo perseguitata dal padre, sì l’uccise.’
Ceffi, Epistole eroiche (1320/30)

iv.

‘Renaldino filiolo dama Avìs molie fu Piero Cristiano di Bari[, deta,] (e) la deta dama Avìs (e) Piero dela Porta, piagi, dieno dare xii li. (e) x s. di p(ro)ve.’
Documento senese
(1263)

  • 23 In modern Italian the diminutive form of a family name has always an expressive or hypocoristic mea (...)

47But if we turn to modern Italian, we observe that, in combination with animate nouns, the diminutive value of -ino has almost completely replaced the old meaning,23 which is preserved only by some animal names (for example: giraffino ‘young giraffe’, leoncino ‘young lion’, etc.).

  • 24 The absence of diminutive outcomes of -īnus in Rumanian suggests that the semantic change came abou (...)

48The absence of Latin words in –īnus with a possible diminutive meaning demonstrates that the semantic shift ‘child/young of X’ > ‘small/little X’ took place quite recently, certainly after the linguistic division of Romània.24

4.2. Greek

  • 25 Data from Buck & Petersen [1944]; translations from Liddell & Scott [1968].

49The most widespread diminutive suffix of Modern Greek, -άκι (cf. ἀνθρωπάκι ‘little man’; γραμματάκι ‘little letter’; κουταλάκι ‘small knife’) is etymologically linked to the Ancient Greek suffix -ίον, which displayed, among other functions, a diminutive value (cf. σκυμνίον ‘little cub’; ταινίον ‘small band’; σαγίον ‘small cloak’).25

50According to Chantraine [1933: 64], the different semantic readings of Ancient Greek words ending in -ίον can be satisfactorily traced back to the unique paraphrase “ce qui appartient à la catégorie de…”, which he considers to be the starting point of the semantic shift that led to the diminutive value, since “ce qui ressemble à une chose peut lui être inférieur, ou être plus petit”.

  • 26 See also Zubin & Köpcke [1986]: “the sex-associated genders are used to identify fully differentiat (...)

51Once again the very first alteration of the original relational meaning has plausibly been triggered by the occurrence of the suffix in combination with animate nouns, that is to designate living beings that are smaller (therefore similar) to the prototype of their species just because they are younger than it. In this case, neuter gender also plays a role: “c’est surtout le genre inanimé du suffixe qui favorisait cette nouvelle évolution. Un nom neutre désigne volontiers de petits êtres, considérés comme «une petite chose»” (Chantraine [1933: 64]).26

52It is not easy to find traces of the intermediate stages of this semantic shift in the most ancient texts of Greek literature, since evaluative morphology was practically unacceptable in epic and lyric poetry as well as in tragedies. Nonetheless, if we glance through a list of the very first occurrences of -ίον, we find many animate nouns the interpretation of which may be traced in the paraphrases ‘young X’ and ‘child of X’:

(17)

κόρη
‘girl’

>

κορίον
‘young, little girl’

θυγάτηρ
‘daughter’

>

θυγάτριον
‘young, little daughter or girl’

ἀλώπηξ
‘fox’

>

ἀλωπέκιον
‘fox cub’

δέλφαξ
‘pig’

>

δελφάκιον
‘suckling pig’

53So, as in the case of Romance languages, even in Greek a suffix with a primary relational meaning begins to designate young living beings and then develops a plain diminutive value.

54This evolutional path seems to occur once again in the course which led Ancient Greek to evolve into Modern Greek. A form such ad δελφάκιον is the trait d’union between ‘ancient’ diminutives in ‑ίον and ‘modern’ diminutives in ‑άκι. The internal structure of such a form underwent a process of reanalysis (#δελφάκ+ιον# > #δελφ+άκιον #), which gave rise to a new suffix -άκιον, “half diminutive and half radical” (Jannaris [1897: 292]). When this new suffix spread and overruled the ancient diminutive ‑ίον, the expressive power of which had been weakened by the extensive use, its first function was that “of forming pet names and nicknames” (Jannaris [1897: 293]).

4.3. Slavonic languages

  • 27 Cf. Scr. sèstrica ‘little sister’, Slov. vªtec ‘small garden’, kôzica ‘little goat’.

55The most ancient Slavonic suffixes displaying a diminutive value (-ec,ĭce and –ĭica / ‑ica, masculine, neuter and feminine respectively)27 can all be traced back to the Common Slavonic suffix *-iko-, which Meillet [1965: 361] describes as follows:

on forme des substantifs désignant des personnes, au masculin, au moyen d’un suffixe *-iko-, qui a pris en slave la forme -ĭce-. […] Le suffixe -ĭcĭ- figure souvent dans les noms d’êtres jeunes: agnę […]: agnĭcĭ «agneau»; telę: telĭcĭ «veau»; etc.; on notera la formation mladěnĭcĭ «enfant», de mladŭ «tendre, jeune». A cet emploi se rattache la formation de diminutifs comme gradĭcĭ «bourg», de gradŭ «ville».

56So, Meillet assigns a crucial role to animate nouns in the emergence of the diminutive value of these suffixes. The semantic shift pointed out by Meillet occurred quite early, since in Old Church Slavonic the diminutive meaning is already predominant:

(18)

cęd(o)-ĭce
child-DIM / ’baby’

iměni(je)-ĭce
possession-DIM / ’small possession’

  • 28 Further evidence for this assumption comes from Baltic languages: in Lithuanian, the diminutive suf (...)

57So, to sum up, the hypothesis that diminutive suffixes are diachronically linked to linguistic items used to express the genealogical relation between father and child (in case of human beings) and/or between the adult and the young (in case of animals) seems to be confirmed by the data. The emergence of diminutive suffixes in Romance languages, Greek and Slavonic languages confirm that the semantic scheme ‘child/young of X’ is a plausible archetype for the diminutive value.28

58Therefore, a diachronic survey of diminutive suffixes of the Indo-European languages of Europe gives evidence for the hypothesis presented in § 2: the development of a linguistic state in which stability is clearly prevalent over frequency is the effect of a very general typological tendency. The main diminutive suffixes attested in European branches of Indo-European family are the consequence of a unique evolutional path, as illustrated in the following map:

5. Diminutives and augmentatives in other linguistic families

  • 29 That is in order to exclude the fact that the similarities we observed in the genesis of Indo-Europ (...)

59The history of Indo-European evaluative morphology reveals a complex and intricate plot in which a very general typological tendency and different areal constraints are involved. Of course, in order to confirm the typological nature of the semantic shift ‘child/young of X’ > diminutive,29 it is necessary to a have a look at the situation of non-Indo-European languages. In this section I will briefly reproduce some interesting data of Bantu, Thai, Austronesian and Mon-Khmer languages, originally discussed in Matisoff [1991] (as regards Thai, Austronesian and Mon-Khmer languages) and Creissels [1999] (with regard to Bantu languages).

60i] Sub-Saharan Africa. Quite recently, a wide renewal of Bantu prefixal morphology has been taking place. As far as evaluative affixes are concerned, Creissels [1999] points out two different tendencies. On the one hand, a new diminutive suffix -ana (i.e. Zulu umfana ‘boy’ > umfanyana ‘little boy (endearing)’) is being produced through the grammaticalization of proto-Bantu word *jana ‘child’. The starting point of this grammaticalization path can be traced in the designation of young animals:

  • 30 In Bantu languages noun classes are conventionally indicated by numbers. In these data, the number (...)

(20)

Shona

mbudz(i)-ana
goat:9-DIM / ’kid’

imbg(a)-ana
dog:9-DIM / ’puppy’

hwai
sheep:9

>

hway-ana
sheep:9-DIM / ’lamb’

huku
fowl:9

>

hukw-ana
fowl:9-DIM / ’chick’30

(Fortune [1955: 120])

61On the other hand, the grammaticalization of the proto-Bantu word *kádį ‘woman’ has given rise to a new augmentative suffix -hadi/-kati:

  • 31 The suffix -hadi/-kati was originally used to indicate the female of a species: Sotho pere-hadi ‘ma (...)

(21)

Sotho

monna-hadi
man-AUG / ’big man’

mosadi-hadi
woman-AUG / ’big woman’

Swati

umutsi-kati
tree-AUG / ’big tree’

litje-kati
stone-AUG / ’big stone’

(Creissels [1999: 32])31

62ii] South-East Asia. Matisoff [1991] identifies two different tendencies in the emergence of evaluative morphology attested in non-related languages spoken in South-East Asia.

63The first tendency concerns the development of diminutive affixes from words meaning ‘child’ (Malay anak ‘child’, Thai lûuk ‘child’, Viet. con ‘child’):

(22)

Mal.

anak
DIM

kuntji

key / ’(small) key’

Thai

lûuk-mεεw
DIM-cat / ’kitten’

Viet.

bàn
table

con
DIM / ’small table’

dao
knife

con
DIM / ’small knife’

64Once again, the intermediate stage of this semantic shift is represented by names of young animals:

(23)

Thai

lûuk-mǔu
child-pig / ’piglet’

lûuk-sïa
child-tiger / ’tiger cub’

Viet.

trâu-con
buffalo-child / ’young buffalo’

65The second tendency entails the grammaticalization of words meaning ‘mother’ (Mal. ibu ‘mother’, W.Hm. niam/niag ‘mother’, G.Hm. ‘mother’, Viet. cái ‘mother’) in order to form new augmentative affixes:

(24)

Mal.

ibu
AUG

kota
city / ’a big city, capital’

ibu
UG

sungai
river / ’a big river, main river’

W.Hm.

ib
one

tug
CL

niag
AUG

neeg
person / ’an important person’

ib
one

tug
CL

niag
AUG

nom
chief

loj
big / ’an important chief’

G.Hm.

nā-Ngâo
AUG-boat / ’big boat’
nā-túanė
N
AUG-person / ’big person (either man or woman)’

Viet.

đòn
carrying pole

cài
AUG / ’big carrying-pole’

hòn
island

cài
AUG / ’main, big island’

66Data presented so far can be represented as follows:

(25)

Semantic archetype of diminutives

Semantic archetype of augmentatives

Italo- and Ibero-Romance zones

young / child

agentive / pejorative

collective

relational

Greek area

young / child

agentive / pejorative

collective

Slavonic area

young / child

locative

Sub-Saharan Africa

young / child

woman / female

South-East Asia

young / child

mother

  • 32 The semantic shift ‘young / child diminutive’ is attested also in Ewe and in Cantonese raised-tone (...)

67So, a quick survey of evaluative morphology of some non-Indo-European languages seems to support my previous conclusions. The emergence of diminutives is often the effect of a very general typological matrix, ‘young / child > ‘small/little’, independently of genetic and areal constraints.32 On the contrary, in the development of augmentatives, a frequent but unstable phenomenon, many different areal tendencies do occur. In other words, since a wide typological tendency is missing, when the category ‘augmentativestarts developing each language autonomously uses the simplest manner to express it on formal grounds. In this case, the simplest way is to ‘reproduce’ the formal strategy adopted by neighbouring languages, even if genetically unrelated.

  • 33 We must be very cautious not to over-extend this conclusion: ‘wide’ and ‘general’ are not synonymou (...)

68So, the history of Indo-European evaluative morphology must not be approached in a unitary way. In fact, an investigation of the history of some Indo-European augmentative suffixes reveals that it is especially difficult to single out recurrent and common stages in their evolutional processes: each language or group of languages seems to proceed along a specific path in order to express the meaning ‘big X’ by morphological means. Nonetheless in the genesis of Indo-European augmentative suffixes many traces of convergence among different languages can be found: in the absence of typological tendencies, languages spoken in the same region seem to influence one another and seem to develop this morphological strategy in the same way. On the contrary, in the case of diminutives, a common archetype can be found, even when a wide cross-linguistic perspective is taken into account.33

6. Conclusions: why dynamic typology?

69In conclusion, it is necessary to refer once again to the list presented in (2). As we stated before, the current typological shaping of European languages with regard to evaluative morphology is the result of a quite recent process. In this regard we have pointed out that most languages of the penultimate generation lacked augmentative suffixes.

70If this is the premise, one can wonder whether dynamic typology is really the best framework in order to explain the history of evaluative morphology. In fact, since both ancient and modern European languages display some diminutive suffixes, the history of their evaluative morphology could be easily explained by stating that a new category, augmentatives, was added to the already present category of diminutives.

  • 34 Cf. Wright [19622: 150]: “-chen (Middle Low German -kīn = K + īn where each element is a dim. suffi (...)

71However, if we examine the issue in detail, this point of view has a number of shortcomings. In fact, as we saw in § 1, in the history of evaluative morphology of Indo-European languages of Europe, besides the already mentioned process of innovation, corresponding to the emergence of augmentatives, there seems to have been also a process of renewal, involving diminutives. This process took place in Romance languages, in Slavonic languages and in Greek: suffixes that displayed a diminutive value in the earliest stages of these languages usually do not correspond to present-day diminutive suffixes. Other linguistic groups, Germanic and Celtic above all, seem not to have been affected by this renewal. In German, for example, the most widespread diminutive suffixes, -chen34 and -lein, did not change in the course of ages: they are the product of a complex syncretism between Old High German diminutive suffixes -īn and -ein and well-known Proto-Indo-European diminutive morphemes *-k(o)- and *‑l(o)- (cf. Butler [1971: 50]). As to Celtic languages, the suffix –an, widely used in Old Irish (ferán ‘little man’ < fer ‘man’), is still preserved in forms such as Scottish Gaelic balachan ‘wee laddie’ (< balach ‘boy’; cf. Thurneysen [1946] and Ball and Fife [1993]).

72Therefore, if diminutives are a stable and frequent phenomenon, why did the renewal not involve all their occurrences? The answer to this question may be found by examining the emergence of augmentative suffixes and this justifies the choice of the framework of dynamic typology. My opinion is that the fact that diminutives have been renewed only in languages that have developed augmentative suffixes did not occur by chance. In other words, it has been the emergence of augmentatives that has triggered the renewal of diminutives. So, one can easily expect that the languages that did not undergo this innovation, have preserved their diminutives. This is the case of Germanic and Celtic languages. Thus, we are not dealing with a simple addition of augmentatives to already present diminutives; the renewal triggered by the innovation suggests that a more complex typological shift took place. In the transition from type A to type B, diminutives also underwent relevant changes. What is really intriguing, is that the renewal also involved diminutive prefixes:

  • 35 In Italian (and in many other Indo-European languages) the most widespread evaluative prefixes are (...)

[p]our ce qui concerne les preéfixes évaluatifs de l’italien, ce qui frappe le plus est que cette classe, à la différence des autres, n’a pas toujours été constituée des mêmes éléments, mais au contraire ses membres ont changé beaucoup au cours du siècle. […] En fait, le système de la préfixation de l’italien ancien était fort different de l’actuel, et la plupart des évaluatifs qui existaient dans les premiers siécles ne sont plus productifs aujourd’hui ou ont complètement perdu cette fonction. […]. [L]a classe des préfixes évaluatifs est celle qui a été la plus mobile pendant toute l’histoire de l’italien, et […] elle n’a jamais cessé de s’enrichir de nouveaux éléments en même temps qu’elle en abandonnait d’autres (Montermini [2002: 218-219]).35

73As a result we may safely state that evaluative affixes seem to represent an autonomous micro-system within derivational morphology. In fact, processes that have led to the emergence of present-day augmentative and diminutive suffixes reveal a complex network of mutual relations in which the success or the failure of each linguistic item depend on the outcome of items linked to it. Within this network, the re-building of evaluative morphology, which took place within the slow and complex course which led from ancient to modern Indo-European languages and which brought about the development of a stable and frequent type, is the result of an interaction between a general typological matrix and some areal-specific tendencies.

Top of page

Bibliography

Ambrazas Saulius, “On the development of diminutives in the Baltic languages”, Linguistica Baltica, Vol. 2, 1993: 47-67.

Ball Martin J. & Fife James (eds.), The Celtic Languages, London – New York, Routledge, 1993.

Buck Carl D. & Petersen Walter, A Reverse Index of Greek Nouns and Adjectives, Chicago, University of Chicago Press, 1944.

Butler Jonathan L., Latin -īnus, -īna, -ĭnus and -ĭneus, Berkeley-Los Angeles-London, University of California Press, 1971.

Chantraine Pierre, La formation des noms en grec ancien, Paris, Champion, 1933.

Creissels Denis, “Origine et évolution des diminutifs et augmentatifs dans quelques langues africaines”, Silexicales Vol. 2, 1999: 29-35.

Cuzzolin Pierluigi, “Sull’origine del singolativo in celtico, con particolare riferimento al medio gallese”, Archivio Glottologico Italiano, 1998: 121-149.

Dressler Wolfgang U. & Merlini Barbaresi Lavinia, Morphopragmatics. Diminutives and intensifiers in Italian, German and other languages, Berlin-New York, Mouton De Gruyter, 1994.

Gaide Françoise, Le substantifs masculins latins en …(I)Ō, …(I)ŌNIS, Louvain-Paris, Éditions Peeters, 1988.

Glare Peter G. W. (ed), Oxford Latin Dictionary, Oxford, Oxford University Press, 1982.

Fortune George, An Analytical Grammar of Shona, London-Cape Town-New York, Longmans, Green and Co, 1955.

Grandi Nicola, “Su alcune presunte anomalie della morfologia valutativa: il rapporto con il genere ed il numero”, Archivio Glottologico Italiano, Vol. 1, 2001: 25-56.

Grandi Nicola, Morfologie in contatto. Le costruzioni valutative nelle lingue del Mediterraneo, Milano, Franco Angeli, 2002a.

Grandi Nicola,Development and Spread of Augmentative Suffixes in the Mediterranean Area”, in Ramat Paolo and Stolz Thomas (eds.), Mediterranean Languages, Bochum, Brockmeyer, 2002b: 171-190.

Grandi Nicola, “Matrici tipologiche vs. tendenze areali nel mutamento morfologico. La genesi della morfologia valutativa in prospettiva interlinguistica”, Lingue e linguaggio, 2003: 105-145.

Greenberg Joseph H., “Some Methods of Dynamic Comparison in Linguistics”, in Puhvel Jaan (ed.), Substance and Structure of Language, Berkeley-Los Angeles, University of California Press, 1969: 147-203.

Greenberg Joseph H., “Diachrony, Synchrony, and Language Universals”, in Greenberg Joseph H., Ferguson Charles A. & Moravcsik Edith A. (eds.), Universals of Human Language, vol. 1: Method and Theory, Stanford, Stanford University Press, 1978: 63-91.

Greenberg Joseph H., “The Diachronic Typological Approach”, in Shibatani Masayoshi & Bynon Theodora (eds.), Approaches to Language Typology, Oxford, Clarendon Press, 1995: 145-166.

Haas Mary R., “The expression of the diminutive”, in Smith M. Estellie (ed.), Studies in Linguistics in honor of George L. Trager, The Hague, Mouton, 1972: 148-152.

Hakamies Keino, Étude sur l’origine et l’évolution du diminutif latin et sa survie dans les langues romanes, Helsinki, Suomalaisen Kirjallisuuden Seuran Kirjapainon Oy, 1951.

Heilmann Luigi, Grammatica storica della lingua greca, Torino, S.E.I., 1963 (off-print of Enciclopedia classica, vol. 5: La lingua greca nei mezzi della sua espressione).

Jannaris Antonius N., An Historical Greek Grammar, London, MacMillan and Co., 1897.

Jurafsky Dan, “Universal tendencies in the semantics of the diminutive”, Language, Vol. 72.3, 1996: 533-578.

Kahane Henry & Kahane-Toole Renée, “The Augmentative Feminine in the Romance Languages”, Romance Philology, Vol. 2, 1948-1949: 135-175.

Lazzeroni Romano, “Per la storia dei derivati in -ōn- nelle lingue classiche”, Studi e Saggi Linguistici, Vol. 3, 1963: 1-48.

Liddell Henry G. & Scott Robert, A Greek-English Lexicon, Oxford, Oxford University Press, 1968.

Matisoff James A., “The Mother of All Morphemes: augmentatives and diminutives in areal and universal perspectives”, in Ratliff Martha & Schiller Eric (eds.), Papers from the first annual meeting of the Southeast Asian Languistics Society, Tempe (AZ.), Arizona State Universit, 1991: 293-349.

Meillet Antoine, Le slave commun, Paris, Champion, 1965.

Montermini Fabio, Le système préfixal en Italien contemporain, Ph. D. Diss., University of Paris X and Bologna 2002.

Montermini Fabio, Il lato sinistro della morfologia. La prefissazione in italiano e nelle altre lingue del mondo, Milano, Franco Angeli, 2009.

Nandriş Grigore, Old Church Slavonic Grammar, London, The Athlone Press, 1965.

Taine-Cheikh Catherine, “Morphologie et morphogenèse des diminutifs en zénaga (berbère de Mauritanie)”, in Naït-Zerrad Kamal (ed.), Articles de linguistique berbère. Mémorial Werner Vycichl, Paris, L’Harmattan, 2002.

Rohlfs Gerhard, Grammatica storica della lingua italiana e dei suoi dialetti, 3 vols., vol. 3: Sintassi e formazione delle parole, Torino, Einaudi, 1969.

Thurneysen Rudolf, A Grammar of Old Irish, Dublin, Dublin Institute for Advanced Studies, 1946.

Vaillant André, Grammaire comparée des langues slaves, Tome IV: La formation des noms, Paris, Éditions Klincksieck, 1974.

Volpati Carlo, “Coppie di nomi di due generi”, Lingua Nostra, Vol. 16, 1955: 2-5.

Wright Jospeh, Historical German Grammar, Vol. 1: Phonology, Word-formation and Accidence, Oxford, Oxford University Press, 19622.

Zubin David A. & Köpcke Klaus-Michael, “Gender and Folk Taxonomy: the Indexical Relation between Grammatical and Lexical Categorization”, in Craig Colette (ed.), Noun classes and categorization, Amsterdam, Benjamins, 1986: 139-180.

Top of page

Appendix

Abbreviations

A(djective), Alb(anian), anim(ate), AUG(mentative), Blg (Bulgarian), Cat(alan), CL(assifier), Cz(ech), Pol(ish), DIM(inutive), Eng(lish), F(eminine), Fr(ench), G(reen) Hm(ong), Grm (German), hum(an), INTF (interfix), Ir(ish), It(alian), Lat(vian), Lit(huanian), LOC(ative), M(asculine), M(odern) Gr(eek), Mac(edonian), Mal(ay), N(oun), NTR (neuter) PEJ(orative), Port(uguese), Rum(anian), Rus(sian), Scr (Serbo-Croat), Sct (Scottish Gaelic), Slov(ene), Sp(anish), Srd (Sardinian), V(erb), Viet(namese), W(hite) Hm(ong).

Top of page

Notes

1 In this paper, by the label ‘evaluative morphology’ I will refer only to its descriptive/quantitative side (represented by diminutive and augmentative affixes), whereas the qualitative side (that is pejorative vs. ‘meliorative’ affixes) will not be considered (cf. Grandi [2002a: 27-35] for internal organization of evaluative morphology).

2 To my knowledge, no counterexample to this state of affairs has been found yet.

3 The notation A (> B?) indicates that in French an augmentative suffix (-ard; cf. meulard ‘big millstone’ < meule ‘millstone’) is gaining ground, but its occurrences are still too sporadic to place French in type B. Furthermore, in some dialects spoken in the Southern part of France a few feminine augmentatives are attested (see § 3.1.2). However, the list presented in (2) refers to standard languages and dialectal variations are not taken into account.

4 The question mark indicates that English displays some diminutive suffixes (for example -let: piglet, booklet), but all of them seem to be completely unproductive.

5 As in the case of French, the notation A (> B?) indicates that in Slovene a few augmentative suffixes are probably being grammaticalized, but their occurrences are still too sporadic to place Slovene in type B.

6 According to Greenberg [1995: 152] the combination of the highest stability and the highest frequency produce universality. In fact, diminutives have often been considered as a ‘quasi-universal’.

7 In this section we summarize the picture already presented in Grandi [2002b], to which we refer the reader for details.

8 Latin data are from Lazzeroni [1963] and Gaide [1988]; Ancient Greek data are from Chantraine [1933].

9 As usual, a particular characteristic gives rise to a nickname that afterwards can ‘crystallize’; in this way it no longer designates a single person, but his (or her) entire family.

10 For Latin, cf. Gaide [1988: 126]: “[l]e sème augmentatif (« virtuel ») que présente le suffixe dans ce type est à l’origine d’une évolution fonctionnelle du suffixe du latin aux langues romanes: dans la Romania du Sud après base substantivale le suffixe a généralement une valeur augmentative (de « virtuel » le sème augmentatif est devenu « dénotatif »; la valeur de « caractérisation » s’est effacée)”.

11 Data of Romance languages are from Kahane / Kahane-Toole [1948-1949] and Volpati [1955].

12 As for Italian, cf. also the very widespread couple pennello masc. ‘brush’ / pennellessa fem. ‘flat brush’ and the following dialectal forms: kortella ‘large kitchen knife’ < kurtello ‘knife’ (Pisa, Tuscany); rastela ‘broad rake’ < rasté ‘rake’ (Alessandria, Piedmont), badíla ‘the road’s worker large shovel’ < badíl ‘shovel’ (Milan and Cremona, Lombardy), etc.

13 Dialectal varieties.

14 See Heilmann [1963: 149]: “La desinenza -α del nom. voc. acc. neutro (ζυγά, μέτρα) rappresenta lo sviluppo greco di uno -*∂ i-europeo […] che […] caratterizzava dei collettivi.”

15 We must be very cautious about projecting this synchonic generalization on a diachronic dimension. In other words, it is not to be taken for granted that singulative derives from diminutives or vice versa. For details see Cuzzolin [1998].

16 Nandriş [19652: 88].

17 We can put forward the hypothesis that a possible link between locative and augmentative meaning is represented by place names that designate a place in which something happens in an exaggerated way or in which some ‘entities’ are present in large amounts.

18 On the contrary, as in the case of the semantic change presented in § 3.1.1, a form such as Late Latin căpĭto may be interpreted both in an ‘old’ agentive/pejorative way, (that is ‘big-headed person’), as well as in a ‘new’ augmentative way (that is ‘big head’).

19 For an exhaustive survey of Latin outcomes of Indo-European denominal suffixes *‑ĭno-, *-īno- and *-eyno- / *-oyno- cf. Butler [1971].

20 Data and translations from Oxford Latin Dictionary (cf. Glare [1982]).

21 Cf. also Rohlfs [1969: 412]: “ciò che è meno compiuto è più piccolo”.

22 F. = son.

23 In modern Italian the diminutive form of a family name has always an expressive or hypocoristic meaning.

24 The absence of diminutive outcomes of -īnus in Rumanian suggests that the semantic change came about when the Latin spoken in the Balkans was already an autonomous linguistic system, that is out of the direct influence of Rome. Cf. Grandi [2003] for details on the relative chronology of the events.

25 Data from Buck & Petersen [1944]; translations from Liddell & Scott [1968].

26 See also Zubin & Köpcke [1986]: “the sex-associated genders are used to identify fully differentiated taxa that have concrete imageability including overall shape and specifiable parts, while neut-gender is used for taxa that do not, and are in this sense undifferentiated” (p. 151) and “[m]asc and fem-gender mark the terms for the male and female adult of each species, while neut-gender is assigned to the nonsexspecific generic and juvenile terms” (p. 174).

27 Cf. Scr. sèstrica ‘little sister’, Slov. vªtec ‘small garden’, kôzica ‘little goat’.

28 Further evidence for this assumption comes from Baltic languages: in Lithuanian, the diminutive suffix ‑énas (žmogénas ‘little man’ < žmogùs ‘man’) is very frequent in kinship terminology (brolénas ‘nephew’, but lit. ‘son of the brother’ < brólis ‘brother’; seserénas ‘nephew’, but lit. ‘son of the sister’ < sesuõ ‘sister’) and in nouns designating young animals (gérvénas ‘young crane’ < gérvė ‘crane’; genýnas ‘young woodpecker’ < genys ‘woodpecker’). Cf. Butler [1971: 18] and Ambrazas [1993].

29 That is in order to exclude the fact that the similarities we observed in the genesis of Indo-European diminutives are due to common genetic inheritance.

30 In Bantu languages noun classes are conventionally indicated by numbers. In these data, the number 9 refers to the noun class 9 (which usually contains nouns designating animals).

31 The suffix -hadi/-kati was originally used to indicate the female of a species: Sotho pere-hadi ‘mare’ (< pere ‘stallion’), tau-hadi ‘lioness’ (< tau ‘lion’).

32 The semantic shift ‘young / child > diminutive’ is attested also in Ewe and in Cantonese raised-tone diminutives (Jurafsky [1996: 539]). Among Indo-European languages, it has probably played a role in the emergence of Germanic and Baltic diminutive suffixes too. The most evident difference between the Indo-European languages and the non-Indo-European languages presented above is that in the latter the starting point of the entire process is a free form, that is a word. In the former, the process involves a bound form, that is a suffix, in all of its stages. Undoubtedly, this difference is relevant, but in this paper I am focusing on the semantics of evaluative affixes and in this sense the analogies overcome the differences.

33 We must be very cautious not to over-extend this conclusion: ‘wide’ and ‘general’ are not synonymous of ‘universal’. The semantic shift ‘young / child > diminutive’ is attested in many typologically different linguistic families, but exceptions are not excluded from this generalization. For example, in most Afroasiatic languages diminutives follow an evolutive path in which animate nouns do not play any role. In Berber the most common morphological strategy in diminutive formation is the circumfix t__t (i.e. Kabyle tadart ‘small foot’ < adar ‘foot’), the primary meaning of which (also in a diachronic perspective, cfr. Taine-Cheikh [2002]) is feminine (i.e. Moroccan Berber tafroukht ‘girl’ < afroukh ‘boy’). The same circumfix is widely attested in singulative forms (i.e. Ayt Ayache Berber talxuxt ‘peaches’ (collective) < lxux ‘a peach’). The use of the diminutive circumfix t___t in combination with animate nouns is generally impossible, since the default interpretation is feminine.

34 Cf. Wright [19622: 150]: “-chen (Middle Low German -kīn = K + īn where each element is a dim. suffix, MHG (Middle German dialect) -chin, -chen).”

35 In Italian (and in many other Indo-European languages) the most widespread evaluative prefixes are mini- and micro- (diminutives) and maxi-, mega(lo)- and macro- (augmentatives). Cf. also Montermini [2009].

Top of page

References

Electronic reference

Nicola Grandi, « Renewal and Innovation in the Emergence of Indo-European Evaluative Morphology », Lexis [Online], 6 | 2011, Online since 27 March 2011, connection on 20 February 2019. URL : http://journals.openedition.org/lexis/403 ; DOI : 10.4000/lexis.403

Top of page

About the author

Nicola Grandi

University of Bologna, Italy
nicola.grandi@unibo.it

Top of page
  • Logo Centre d’études linguistiques
  • Logo Université Jean Moulin – Lyon 3
  • OpenEdition Journals