Alexander Murray,
a Scottish clergyman and orientalist
on the origins of language

András Cser

The Author

The purpose of this paper is to present a now little-known work published two hundred years ago by a now little-known author (who was relatively well known at his time). This is Alexander Murray’s Philosophical History of the European Languages, published in 1823.The book is freely available at this URL: books.google.hu/books?id=oJpJAAAAcAAJ The central question around which this book revolves is the emergence and the prehistory of the languages that we now call Indo-European or, more precisely, a subset of these languages (since some of them were not known at all at that time, such as Hittite or Tokharian, some were not known to be Indo-European, such as Armenian or Albanian).

But who was this Alexander Murray (1775–1813)? He was a Scottish clergyman and philologist with a very humble background, who rose to some prominence in 1812, when he was elected professor of Oriental philology at Edinburgh University. Before he actually started teaching he published a grammar of Hebrew called Outlines of Oriental Philology, and then after just a couple of lectures his health broke down and he died of consumption at the age of 38. The book called Philosophical History of the European Languages was published posthumously ten years after his death, and was edited by fellow orientalist David Scot. The book is based on Murray’s extensive handwritten notes, but the final arrangement (though not the wording itself) is Scot’s work. The book became quite well known, and only two years after its publication it was even translated into German.

The Philosophical History of the European Languages

The structure of the book is the following. A lengthy editorial preface by David Scot explains how the book came into being and expatiates upon Murray’s professional carreer. Then a ninety-page long biography of the author follows. The core of the book is the discussion — only about 140 pages long — divided into six chapters (see original Table of Contents below).

figure 1
The table of contents of the Philosophical History of the European Languages

The discussion is accompanied by more than 300 pages of notes called Facts and illustrations, and this is where Murray relegated most of his data, the scant references to other authors, and some of his more extended arguments.

The central part of his arguments is found in chapter 3 (Origin of the European languages), and this is what we will focus on here. Murray made up a theory of how these languages came to be, more precisely, how the ancestor of these languages came to be and how it developed later. The idea is that originally there were only nine monosyllables all ending in AG (AG, BAG, DWAG…) and variants thereof. These captured elementary and very general meanings and sufficed for communication at a very elementary stage in the history of mankind. After a while, however, these monosyllables were combined with each other, some functioning as semantically central elements, some as modifying elements. The complex forms created at stage two then gradually developed into the forms found in the documented languages (with data mainly taken by Murray from Germanic languages, Greek and Latin). The following passage from the first pages of chapter 3 explains the first stage.

Taste and philosophy will receive with aversion the rude syllables, which are the base of that medium, through which Homer, and Milton, and Newton, have delighted or illumined mankind. The words themselves, though inelegant, are not numerous: each of them is a verb and name for a species of action. Power, motion, force, ideas united in every untutored mind, are implied in them all. The variation of force in degree was not designated by a different word, but by a slight change in the pronunciation. Harsh and violent action, which affected the senses, was expressed by harsher articulations.
I. To strike or move with swift equable penetrating or sharp effect was AG! AG! If the motion was less sudden, but of the same species, WAG. If made with force and a great effort, HWAG…
II. To strike with a quick, vigorous, impelling force, BAG or BWAG, of which FAG and PAG are softer varieties…
IX. To move with a weighty strong impulse, SWAG.
These nine words are the foundations of language, on which an edifice has been erected of a more useful and wonderful kind, than any which have exercised human ingenuity. (pp. 31–32)

A few pages later Murray explains that certain languages, such as Chinese, are stuck in a sense at this stage, and make use of subsidiary measures, but others have moved on:

The Chinese, whose language continues to be monosyllabic, had recourse to the expedient of varying the sound with the sense, a method sufficient to serve ordinary purposes, but of narrow compass, and liable to difficulties in practice. But the fathers of those nations, whose languages were to receive the most abstract or animated thoughts which the mind is capable of forming, began early to compound their words, and to multiply terms with all the fertility of arithmetical permutation. (p. 34)

And then he goes on to explain the workings of language at the second stage, when these primitive elements combine with one another:

These words, of which the general and particular applications were familiar to every individual, when annexed to one another, modified the proper meaning of each radical, altered its sense from an absolute to a limited state, and expressed circumstances of time, degree, and manner of action. An example will illustrate this part of the subject. The radical WAG, as has been stated, signifies to move, shake, or agitate. This is its original unrestricted sense, not limited by time or any other circumstance. When GA, go, or DA, do, are joined to it; WAGIDA, which is a contraction for WAG-DAG, expresses that the action is finished or done; and GAWAGIDA, that it is done and gone by. This is the origin of the imperfectly preterite and perfectly preterite tense and participle in all the Teutonic dialects. (p. 36)

Or, to take a more complex case:

Moderation is, in all its parts, MOG-DA-RA-TI-GA-NA-GA, formed in this succession; MAG, seize, comprehend, include, contain, measure; MOG-DA, measured, the preterite participle by DA, done: whence MOD and SA; MO­DUS, measure, bound; and SA agency, which is implied in all ancient nouns. Add RA, work, to MOD; there results MODERA, was making to have bounds, keeping in bounds; whence MODERATA, a preterite participle, kept in bounds. Add IG, make, to MODERA; and you have MODERATIG, an adjective, which means making kept in bounds, or having the quality of being kept in bounds. To MODERATIG join ANGA or ONGA, a compound of NA, make, and GA, go, which is the origin of our ING in present participles; and MODERATIGONGA is obtained, an abstract noun quite analogous to the Teutonic; BEWEGUNG, motion; HILDIGUNG, inclining; ERMAHNUNG, admonition. (p. 196)

The Background

Having briefly looked at the work itself and some of the central issues it addresses, let us turn to the question what intellectual and linguistic traditions Murray continues in his discussion.

Three such traditions can be discerned, which may be loosely termed philosophical, philological and grammatical. Of these the philosophical is the oldest and the most varied (not surprisingly), and itself includes two different directions of inquiry. One concerns the emergence of language, a highly popular topic in the two hundred years preceding Murray’s time. It is well known that in the seventeenth and the eighteenth centuries respectively two rather different views dominated regarding the ultimate origins of language (Aarsleff 2006). In modern terms, the issue was whether language originated in the cognitive or the social functions of human beings. The latter view was more typical of the eighteenth century and usually entailed a greater role played by emotions, gestures and imitation in the emergence of language. Interestingly, Murray does not explicitly quote the English and Scottish philosophers who had expounded on the subject (such as Locke or Lord Monboddo); the only such philosopher he refers to (though not by name) is Adam Smith (“this is the opinion of the illustrious author of the Theory of Moral Sentiments and the Essay on the First Formation of Language” p. 178).

The other originally philosophical issue was the composition of words and, coupled with that, the question of arbitrariness. There is a time-honoured tradition going back to Plato that seeks to explain the meanings and origins of words with reference to their constituent sounds/letters and/or with reference to contracted phrases (cf. Plato’s explanation of anthropos ‘man’ as anathron ha opope ‘who looks up at what he has seen’, early medieval explanations of, e.g., petra ‘stone’ as pedibus trita ‘trodden on by feet’). Contraction as a working method of etymology was carried to extremes by the Englishman Horne Tooke in his Diversions of Purley (published in Murray’s lifetime, between 1786–1805), and it is clear that Horne Tooke had a great impact on Murray, as indeed on other contemporaries too, who compared the work of both of them to the achievements of chemistry, then the par excellence science.

The philological tradition in this case means the philology of the Se­mit­ic languages on the one hand, with which Murray was familiar as an orientalist and which, by his time, had a European tradition going back three hundred years. On the other hand, it meant the recent interest in the European vernaculars and their history, including the numerous grammars of English published over the 18th century, Samuel Johnson’s dictionary and the first editions of Old English texts. Specifically, Murray is greatly indebted to George Hickes’s Linguarum veterum septentrionalium thesauri grammatico-critici & archæologici, a work on what they called northern languages and what we call Germanic languages now.

The grammatical tradition in this case is that of the analysis of word structure. As is well known, in the history of European grammatical thinking, words were not decomposed into morphemes until the 16th century, when Hebrew began to have its impact on grammatical analysis. We cannot here go into the full story of word analysis over several centuries, but the outlines are the following. There were, after 1500, two relevant notions, that of root/stem and theme. The former meant the formal core of a word, in a fashion somewhat similar to the modern use of the term. The latter meant the starting point of a paradigm, essentially the first item in a list of forms, with nouns always the singular nominative, with verbs usually the first person singular present tense (sometimes the third person singular, especially in Hungarian grammar). The two terms began to converge quite early on and in many grammars they appear as synonyms, as seen, e.g., in Palsgrave (1530: xxxi): “His [=a verb’s] thre chefe rotes, that is to say, his theme, his preterit participle and his present infynityve… je parle, jay parlé, parler.

When the notion of root became established in the European grammatical tradition, it was a purely practical descriptive device. Soon, however, it led to interesting questions that subsequently ramified into some of the most interesting language-related questions of the period. One such question was whether roots historically precede their derivatives, that is, are older languages more likely to exhibit more roots and fewer derivatives than languages at a later stage of development?

Before the nineteenth century, this question does not really emerge in this form for the reason that the notion of empirical historicity as opposed to abstract notional derivations was poorly understood (see Telegdi 1967). One exceptionally explicit discussion of sorts is found in works of Mäzke, a German grammarian of the late eighteenth century (e.g., 1776, 1780). He makes a distinction between two stem-notions, the Grundsilbe and the Stammwort. The former means the stem in the practical-descriptive sense, i.e., what remains of a word form when all affixes have been removed. The latter means the ultimate core of the word, from which the Grundsilbe as well as its whole family of forms can be derived. For instance, the word erröten ‘redden’ has the Grundsilbe röt but the Stammwort rot (‘red’). Mäzke cosistently uses the notion of Grundsilbe in a synchronic sense, while Stammwort also receives a diachronic interpretation. This, in a way, is an anticipation of the strongly diachronically oriented hypotheses of roots that so strongly defined linguistics in the nineteenth century.

Another question is how affixes relate to roots: did they develop from original autonomous roots themselves (as in what is now called grammaticalisation), or on the contrary, are they “growths” on their host roots that “sprout” out of the latter, as it were? Of the two possible answers the former has a long pedigree ultimately harking back to Aristotle and to modistic grammar, followed by Port Royal among others in the claim that all verbs include the verb meaning ‘to be’. Also, this view was supported by the observation that in some languages personal pronouns and personal endings on verbs correspond formally as well as functionally (Arabic anta ‘you-masc’ ~ faʿalta ‘you did-masc’ vs. anti ‘you-fem’ ~ faʿalti ‘you did-fem’ etc.). Already in the early eighteenth century some German grammarians described suffixes such as -heit, -tum or -lich as deriving from full words. In the early nineteenth century, this idea developed into a full-fledged theory of Indo-European historical morphology in the hands of Franz Bopp (1816, 1833–52), and while many of the particular arguments made at that time no longer hold, it is still true that contemporary theories of grammaticalisation are the intellectual descendants of these ideas.

The latter answer to this question also has a long pedigree. Already the Renaissance scholar Justus Cæsar Scaliger claims (1540) that the earliest forms of language had no morphology at all, and several early modern scholars claimed the same for the earliest, unattested, stages of particular languages (e.g., Vossius 1635 for Greek). By the eighteenth century theories of monosyllabic primeval languages abound all over Europe and also find their way into the root theory of Schlegel (1808) and early Indo-European linguistics.

A third question, inextricably interwoven with the above two, also emerges if one looks at the development of linguistic thought in the 17th–19th centuries. This is a position rather than a question really: the idea very quickly gained ground in this period that roots are not necessarily attested entities (as they were in the Semitic tradition from which the notion of root comes); they may be highly abstract linguistic units that hardly resemble any actual form found in the language for which they are claimed. Since there were no methodological worries about positing such roots, these theories soon began to blossom and became a dominant trend on the continent (grammarians such as Philipp Zesen, for instance, in the mid-seventeenth century would derive all words from CV-roots from a set of four C-s and four V-s, see Jellinek 1913–1914).

It is quite clear that Murray was part of an intellectual climate that may be regarded as dominant in his time (regardless of how many 17th–18th-century German grammars he had actually read — we certainly do not know). It is also clear that this is precisely the reason why towards the end of the 19th century his work fell into oblivion. By that time the Neogrammarian approach, which was methodologically very strict, and required phonological systematicity for any etymological claim to hold, became dominant and had no real competition from any other approach. And thus scholars like Murray were relegated to the fringes of history and are by now the objects of antiquarian, rather than linguistic, interest.


Aarsleff, Hans. 2008. Philosophy of Language. In: Knud Haakonssen (ed.), The Cambridge History of Eighteenth-Century Philosophy. Cambridge: Cambridge University Press. 451–495.

Bopp, Franz. 1816. Über das Conjugationssystem der Sanskritsprache in Vergleich mit jenem der griechischen, lateinischen, persischen und germanischen Sprache. Frankfurt a. M.: Windischmann.

Bopp, Franz. 1833–52. Vergleichende Grammatik des Sanskrit, Zend, Griechischen, Lateinischen, Lithauischen, Altslawischen, Gotischen und Deutschen. Berlin: Dümmler.

Jellinek, Max Hermann. 1913–1914. Geschichte der neuhochdeutschen Grammatik. 2 vols. Heidelberg: Winter.

Mäzke, Abraham Gotthelf. 1776. Grammatische Abhandlungen über die Deutsche Sprache. Breßlau.

Mäzke, Abraham Gotthelf. 1780. Über Deutsche WörterFamilien und Rechtschreibung. Züllichau.

Palsgrave, John. 1530. L’éclaircissement de la langue françoise. London.

Scaliger, Justus Cæsar. 1540. De causis linguæ latinæ. Lyon.

Schlegel, Friedrich von. 1808. Über die Sprache und Weisheit der Indier. Heidelberg: Mohr & Zimmer.

Telegdi, Zsigmond. 1967. Struktur und Geschichte: zur Auffassung ihres Verhältnisses in der Sprachwissenschaft. Acta Linguistica Hungarica 17: 223–243.

Vossius, Gerard Johannes. 1635. De arte grammatica. Amsterdam.