The panel will focus on the answers provided by different traditions to the question: How does a language work? Various – possibly all – linguistic speculations have dealt with the problem of language representation, trying in particular to spot the components of the language and to understand the way these interact: each investigation on language has in fact dealt with concepts such as sound, word, sentence, etc., or other similar devices in order to represent the functioning of languages.
On the other hand, one could ask whether those different traditions, apparently sharing the same aim, are actually mutually understandable. For instance, phonology, morphology, syntax, etc. are grammatical categories developed -through a long and complex history- by Western linguistics, but do they find counterparts in other linguistic speculations, such as the Islamic ones, the South and Far East Asian ones, or their predecessors, i.e. the Greek and Latin ones?
In this respect, South Asia represents a remarkable case since the grammatical models developed for Sanskrit have been applied to various languages such as Pali, Tamil, Tibetan, Old Javanese, etc. and have been as well employed in the formulation of various strands of the Western linguistic speculation: starting from this case study, it will be possible to extend the discussion to other traditions.
Chairperson: Giovanni Ciotti
Presentation of individual papers
Cristian Pallone (Università di Roma – La Sapienza)
Motoori Norinaga’s view on classical Japanese vocalism
Among the philological studies of classical literature and historiography, Motoori Norinaga (1730-1801) developed a complete analysis of the Sino-Japanese readings of Chinese characters, which led him to undertake a more general reflection concerning the sounds of Japanese language. Norinaga wrote about Japanese phonology, vocalism in particular, in the theoretical prefaces of two works “Kana usage for the reading of characters” (Mojigoe no kanazukai, 1776) and “Reflections on the three readings of Chinese characters” (Kanji san’onkō, 1785). In his studies, the philologist borrowed some descriptive categories from traditional siddhamic, sinological and musicological studies and used them innovatively. In particular, through a re-definition of the categories of ‘open-close’ and ‘light-heavy’, Norinaga advanced a descriptive hypothesis on classical Japanese vocalism that has to be understood as based on the shape of the oral cavity in the process of articulation, rather than based on the position of lips or on the pitch modulations, as previous studies have asserted (Kuginuki, 1998 and Matsushige, 2002, 2008).pdf
Philomen Probert (University of Oxford)
Underlying forms and derivations in ancient Greek theory of prosody
Modern linguistic theory often makes use of underlying forms and derivations in the description of complex patterns of alternation. We do not normally associate the use of such tools with ancient Greek scholarship, yet a look at ancient Greek theory of prosody reveals several, such as an underlying acute accent that may be ‘put to sleep’ in connected speech, and even ‘woken up’ again. This paper will illustrate the use of such concepts in ancient Greek theory of prosody, and will argue that some surface features of the language, such as the distinction between long and short vowels, came to be treated as underlying features when they disappeared from the surface phonology. I will also suggest that Greek theory of prosody would have looked very different if other choices had been made.pdf
Stefano Seminara (Pontificio Istituto Biblico – Rome)
‘Linguistics’ in ancient Mesopotamia/span>
Starting from the 2nd millennium B.C., i.e. from the beginning of the period known as Old Babylonian, the Babylonian culture is characterised by a specific variant of bilingualism, known as “literary bilingualism”: the spoken language (spoken and written in the practical and every day use) is Akkadian, whereas the literary language is – together with Akkadian – Sumerian, i.e. the “dead” language of the most ancient Mesopotamian civilization. The ancient works of the Sumerian literature were first translated into Akkadian during the Old Babylonian period. The Babylonians did not leave a ‘manual of linguistics’. The translations of literary works from Sumerian into Akkadian are, together with the so called Grammatical Texts, the only fields in which a Babylonian linguistics developed. According to the bilingual principle of the Mesopotamian culture of the 2nd millennium, the Grammatical Texts are long lists of Sumerian grammatical forms with their Akkadian translations, sometimes followed by grammatical explanations, whose meaning remains rather obscure. More interesting from a linguistic point of view are the literary translations, in which the theory had to be put into practice. From this practice a system of rules was set up and this knowledge was entrusted to a sort of manual of translation (the so-called Examentext A). From all this material emerges an organic idea of language, very far from our traditional concepts, such as semantics and syntax: that’s why a Sumerian word could be translated into its Akkadian antonym and in spite of the great differences between Sumerian and Akkadian (the former agglutinating and ergative, the latter flexional and accusative) the two languages were always regarded and treated as perfectly symmetrical systems. The reason is that the Babylonian interest in linguistics was quite different from our own.
Maria Piera Candotti (Université de Lausanne); Tiziana Pontillo (University of Cagliari)
Underlying forms and derivations in ancient Greek theory of prosody
The question on the existence of different hierarchical layers in Pāṇini’s grammar is an oft debated one, and one in which the influence of different ideological and cultural backgrounds is strongly felt. The kick-off of this debate may well be the well-known article of Kirparsky and Staal in 1969 where the two authors proposed a hierarchy of four levels of representation with a definite Chomskian touch. The discussion has continued till our days. In our communication we will focus on the commonly held assumption of inherently discreet layers and on its operational consequences. In particular we will try to prove at least two points:
- That there is no fixed status for the linguistic items in grammar. One and the same linguistic item can be treated like a phone or like a morph following the derivational needs. It is thus not advisable to identify a given layer with a class of elements.
- That the distinction between what’s morphological and what’s phonological is not sharp cut: in fact there is a kind of continuum going from prototypically phonological items to prototypically morphological ones.
These two points are not intended as a confutation of the existence of layers structuring the Aṣṭādhyāyī but as a suggestion for further reflection on this concept.pdf
Round table: “Borrowing representational devices across language speculation” (provisional title), i.e. What does happen when representational devices developed by a tradition to describe a language A are employed to describe a language B?
The round table will be introduced by the following presentation:
-A Contribution to the History of the Concept of Root (Dr. Luca Alfieri, Università di Roma – La Sapienza)
The term root is usually defined as the input form in the derivational processes. National, say, is the root on which nationalize is built. Yet since national is not a primary item, one may object that the root is broadly defined as the base form for derivational processes but, in a strict sense, refers only to the primary item used as an input for such processes (i.e. nation). In this meaning, the root is a lexical unit and a word-form unit (apart from inflections), as it is stored in the lexicon but can also be used as a word directly (apart from inflection). The term root, however, may refer to a different concept, namely the minimum lexical item of languages such as Vedic, Avestan and, more generally, the ancient Indo-European and Semitic languages. In this case, the root is a primary item, is the input for the derivational processes and is stored in the lexicon but cannot be used as a word directly. Even if inflections are disregarded, the root needs processing through at least one derivational morpheme before it can be used in real sentences. Two meanings are thus hidden under the single label of root: root1 ‘primary item that is also a word-form (apart from inflection)’ vs. root2 ‘primary item that is not a word-form’. These meanings are customarily merged into the single label root, as if being a primary item was more fundamental than being or being not used as a word directly for the theory of grammar, but it is not fully clear what theoretical reasoning determines such a view. This paper aims to argue that, contrarily to what is usually assumed, the two meanings must be separated and only the latter (‘primary item that is not a word-form’) is rightly termed as root. A theoretical-historical survey of various ways in which the notions of root1 and root2 have been conceived throughout the 19th and the 20th centuries in the most commonly employed frameworks of language description is presented in order to demonstrate this view.