Loanword

Translations

Prestito linguistico | emprunt lexical | Lehnwort

Article

Form of lexical interference. It involves the transfer of a linguistic sign (typically a word) from a model to a target language, in which the signifiant of the model is reproduced into the target, with or without a certain degree of phonological adaptation and with or without morphological adaptation, and the meaning may also undergo more or less drastic changes (as a general rule, a word is usually not borrowed with all its range of meanings).

Borrowing occurs in contexts of either direct or indirect cultural contact, and may involve languages that belong to close or distant areal settings. Contrary to some hyper-functionalist interpretations, no functionalist or emic reason (e.g. lack of a designation in target language) is strictly required for the borrowing of words to occur. Structural constraint, on the other hand, may apply, based on types of words that are more or less easily transferred (see borrowability scale).

A traditional distinction is made between adapted and non-adapted loanwords, the former including loans whose signifiant is adapted to the phonological rules of the target language and are assigned its inflectional morphology, thus becoming not immediately recognisable as borrowings (this is particularly evident in the Italian word lanzichenecco, borrowed from German Landsknecht, in which the original complex consonant clusters are simplified according to the Italian phonology and the word also receives an inflectional morpheme in final position), while the latter includes those loans whose original significant remains (more or less) unchanged in the target language, and thus are clearly identifiable as foreign words.

Note that phonological integration should be distinguished from phonetic adaptation, which almost always occurs in the target language (Italian mister is commonly pronounced as ['mistɛr], vs. its English model ['mɪstə]). On the other hand, morphological integration only concerns overt morphology, because the acquisition of some morphological features of the target language is unavoidable (e.g., in a target language with a pervasive gender system, borrowed words, even if non-adapted, are necessarily assigned to a grammatical gender, regardless of the addition of a specific morpheme).

To the basic distinction between adapted and non-adapted loanwords, some linguists add a halfway type, the so-called “acclimatised” loanwords (see e.g. Cotticelli Kurras 2012, with further references), i.e. phonologically and morphologically non-adapted loanwords that, however, are well established in the lexicon and produce derivative words based on productive word-formation rules of the target language (this is the situation of many loanwords in Italian, e.g. bar, scanner, etc., from which the noun barista and the verb scannerizzare are derived respectively). However, base productivity should be regarded as a distinct parameter, because adapted loanwords may or may not be productive too. Therefore, based on the two criteria of phonological and morphological integration and base productivity, the following categorization in four types established by S. Merlin and V. Pisaniello is employed in our framework:

Type of loanword	Phonology and inflectional morphology of the target language	Base productivity in the target language
I = non-productive foreign word	-	-
II = productive foreign word	-	+
III = non-productive integrated loanword	+	-
IV = productive integrated loanword	+	+

Which can be summarized as a two-coordinate system as follows:

	- base productivity	+ base productivity
- phonology and morphology	Type I	Type II
+ phonology and morphology	Type III	Type IV

This categorization should not be understood as strictly hierarchical. A loanword entering the target language may immediately be phonologically and morphologically integrated e become productive. Furthermore, different types for the same loanword may coexist, but sometimes a diachronic distribution can be detected (Hitt. zippulašne ‘bread’ [< Hattian] is “neuter” in OH [possibly a non-adapted form – i.e. type I – but no unambiguous cases], and it is later assigned to the common gender i-stems and inflected accordingly [type III]).

It is sometimes difficult to evaluate the productivity of the base in a clear way: Lyc. *xssaθrapaza- ‘satrap’ (< Iran. *xšaθra-pā-, adapted with the Lycian “professional” suffix -aza-) can be defined as productive since it produces the verb xssaθrapaza- ‘rule as satrap’, but can Lyc. xssadrapa- ‘satrap’ (borrowed from the same Iranian base, but without adding the suffix -aza-) be defined as productive only because of the relational adjective xssadrapahi (in a language in which the use of the relational adjective is a strategy equivalent to the use of the genitive case)? It also should be stressed that base productivity strictly concerns word-formation rules of the target language, and should not be confused with the possible presence of derivatives of a given loanword that are independently borrowed and may only point to the productivity of the base in the model language (cf. e.g. the various Hittite borrowings based on the Hurrian root šehl- ‘pure’, the adj. šehelli- ‘pure’, the nouns šehelli/a- ‘purification (ritual/festival)’ and šehelliški- ‘purification ritual’, all independently borrowed from Hurrian, and the noun šehellit(a)- ‘purification vessel(?)’, possibly borrowed through Luwian intermediation).

To sum up, only phonological and morphological integration are relevant to strictly define a loanword as integrated, while base productivity is a distinct criterion, with no correlation with the degree of integration (note that many native bases may be unproductive in a language).

	- base productivity	+ base productivity
- phonology and morphology	Type I	Type II
+ phonology and morphology	Type III	Type IV

When dealing with loanwords, one should also take into account the path of transmission, because the ultimate source of a borrowing may not correspond to the model language from which it is transferred to the target language. In other words, borrowing processes may be direct or involve intermediate stage of transmission through other languages (when this occurs in a large area of diffusion, one is dealing with a Wanderwort). As far as Anatolian languages are concerned, e.g., several Akkadian words are borrowed in Hittite via Hurrian intermediation, which is made evident by the presence of Hurrian morphemes, pointing to full integration of such loanwords in Hurrian (cf. e.g. Hitt. pūhugari- ‘replacement, substitute’ < Hurr. pūhugari- ‘id.’ < Akkad. pūhum ‘id.’, adapted with the addition of the Hurrian derivative suffix -ugar-).

Sometimes, multiple paths of transmission can be identified for the same word, as in the case of some Hurrian nouns – either originally Hurrian or borrowed from Akkadian – that enter Hittite through two different paths: on the one hand, a direct Hurrian > Hittite path, which usually involves their integration as common gender nouns in Hittite; on the other hand, an indirect path with a Luwian stage, where these nouns are generally adapted as neuter stems in -it, which are then transferred as such in Hittite (cf. Carruba 1967, Giorgieri 2012, Pisaniello 2017).

Examples

For ancient languages, the allocation of each loanword to one of the four types can only be based on currently available data (i.e. the classification of loanwords into types I-III is always provisional, while a type IV loanword is generally unproblematic, although a preceding or coexisting and concurrent loan of different type could exist). The Anatolian and Ancient Near Eastern areas are no exception to this rule. Examples, however, are manifold; here we offer some examples based on the data on pre-classical Anatolia, mostly from Hittite, whose large corpus allows a better evaluation of the different cases.

A general tendency to assign its own inflectional morphology sometimes makes the identification of non-integrated loanwords rather difficult in Hittite. However, some Luwian loanwords are currently only attested as foreign words in Hittite texts, i.e. with unique Luwian ending, e.g. the verb tapar- ‘to rule’, which is an example of type I loan (vs. related tapariya- ‘id.’, also borrowed in Hittite, but morphologically integrated and productive, i.e. a type IV loan). In Bronze age Luwian, non-integrated loanwords are virtually non-existent, given a marked tendency to adapt borrowings as neuter stems in -it (see e.g. Hurrian loanwords in Luwian collected by Simon 2020).

As mentioned, the same loanword may at times occur both as a foreign word and as an integrated loan, e.g. šinapši- (a building, from Hurrian), sometimes occurring uninflected with genitive and dative function, besides the corresponding integrated forms šinapšiaš and šinapšiya (cf. CHD Š, 375-378), and perhaps zippulašne ‘bread’ (from Hattian), seemingly uninflected in OH (but only with nominative and accusative function, thus an interpretation as a neuter integrated loan cannot be excluded) and later occurring as a common gender i-stem. No derivatives are attested so far in Hittite; therefore, these cases illustrate both type I and type III.

Examples of type IV loanwords can be easily found in the Anatolian corpora. An interesting case is the Hitt. noun arzana-, denoting a building, usually translated as ‘inn, hostel, brothel’: it is a loan from Akkadian arsānu, designating a kind of groats, and was borrowed and integrated in Hittite with its original meaning, since it occurs in genitive case determining the noun per/parn- ‘house’ (Sum. É), in a phrase literally meaning ‘porridge-house’ (arzanaš per/parn-). Later, arzana- probably came to metonymically designate the building itself, because we find examples like acc.sg. É arzanan and abl. É arzanaz, where É should be regarded as a determinative, and the word became productive with this new meaning in the Hittite lexicon, producing the derivative ^LÚarzanala- ‘attendant of the arzana-house’ and the denominative verb arzanai- ‘quarter, billet’ (cf. Yakubovich 2006, 44-45).

References

Carruba, O. (1967), Über die “churritischen” Deklinationsformen im Hethitischen, Revue hittite et asianique 25, pp. 151-156. CHD = Güterbock, H.G., Hoffner, H.A., van den Hout, Th.P.J. (eds., 1980-), The Hittite Dictionary of the Oriental Institute of the University of Chicago, Chicago: The Oriental Institute of the University of Chicago. Cotticelli Kurras, P. (2012), Integrazione lessicale e categorie morfologiche dei prestiti luvi in ittito, in P. Cotticelli Kurras, M. Giorgieri, C. Mora, A. Rizza, F. Giusfredi (eds.), Interferenze linguistiche e contatti culturali in Anatolia tra II e I millennio a.C. Studi in onore di Onofrio Carruba in occasione del suo 80° compleanno (Studia Mediterranea 24), Genova: Italian University Press, pp. 73-85. Giorgieri, M. (2012), Anaḫi, anaḫiti: luvio o hurrico?, in P. Cotticelli Kurras, M. Giorgieri, C. Mora, A. Rizza, F. Giusfredi (eds.), Interferenze linguistiche e contatti culturali in Anatolia tra II e I millennio a.C. Studi in onore di Onofrio Carruba in occasione del suo 80° compleanno (Studia Mediterranea 24), Genova: Italian University Press, pp. 139-152. Pisaniello, V. (2017), Hittite (NINDA) kaz(za)mi(t)-, Vicino Oriente 21, pp. 71-82. Simon, Zs. (2020), Die hurritischen Lehnwörter im Keilschriftluwischen, in Romain Garnier (ed.), Loanwords and Substrata in Indo-European Languages. Proceedings of the Colloquium held in Limoges (5th ‒ 7th June, 2018) (Innsbrucker Beiträge zur Sprachwissenschaft 164), Innsbruck: Institut für Sprachwissenschaft der Universität Innsbruck, pp. 411-426. Yakubovich, I. (2006), The Free-Standing Genitive and Hypostasis in Hittite, Journal of Near Eastern Studies 65, pp. 39-50.