Samizdat

"Samizdat: Publishing the Forbidden."

Tag: University of Oxford

Eco: Francis Lodwick

41_full

Francis Lodwick (1619-1694), eight verses from the first chapter of the Gospel of St. John in Francis Lodwick’s common writing, next to the numerical key composed for it. A Common Writing, London, 1647. Museum of the History of Science, University of Oxford. This work is in the public domain in its country of origin and other countries and areas where the copyright term is the author’s life plus 100 years or less.  

Lodwick wrote before either Dalgarno or Wilkins, both of whom had thus the opportunity to know his work. Salmon (1972: 3) defines him as the author of the first attempt to construct a language in universal character. His first work, A Common Writing, appeared in 1647; The Groundwork or Foundation Laid (or So Intended) for the Framing of a New Perfect Language and a Universal Common Writing dates from 1652.

Lodwick was not a learned man–no more than a merchant, as he humbly confessed. Though, in his Ars signorum, Dalgarno praised Lodwick for his endeavors, he was unable to hold back the supercilious observation that he did not possess the force adequate so such an undertaking, being a man of the arts, born outside of the Schools (p. 79).

In his writings, Lodwick advanced a number of proposals, some more fruitful than others, on how to delineate a language that would both facilitate commercial exchange and permit the easy acquisition of English.

His ideas, moreover, changed over time, and he never managed to design a complete system. None the less, certain of what appears in the most original of his works (A Common Writing, hardly thirty pages long) reveals him as striking off in a direction very different from other authors of his time, making him a precursor of certain trends of contemporary lexical semantics.

In theory, Lodwick’s project envisioned the creation of a series of three numbered indexes; the purpose of these was to refer English words to the character and these to its words. What distinguished Lodwick’s conception from those of the polygraphers, however, was the nature of its lexicon.

Umberto Eco, The Search for the Perfect Language, Figure 13.1, p. 261

Umberto Eco, The Search for the Perfect Language, Figure 13.1, p. 261.

Lodwick’s idea was to reduce the number of terms contained in the indexes by deriving as many of them as possible from a finite number of primitives which express actions. Figure 13.1 shows how Lodwick chooses a conventional character (a sort of Greek delta) to express the action of drinking; then, by adding to this radix different grammatical marks, makes the different composite characters express ideas such as the actor (he who drinks), the act, the object (that which is drunk), the inclination (the drunkard), the abstraction, and the place (the drinking house, or tavern).

From the time of Aristotle up until Lodwick’s own day, names of substances had invariably been the basis upon which a structure of classification had been erected. Lodwick’s original contribution, however, was to commence not with substantives but with verbs, or schemes of actions, and to populate these schemes with roles–what we would now call actants–such as agent, object, place and so on.

Lodwick designed his characters to be easily recognized and remembered: as we have seen, to drink was specified by a sort of Greek delta, while to love was a sort of L. The punctuation and added notes are vaguely reminiscent of Hebrew. Finally, as Salmon suggests, Lodwick probably took from contemporary algebra the idea of substituting letters for numbers.

In order to set up his finite packet of radicals Lodwick devised a philosophical grammar in which even grammatical categories expressed semantic relations. Derivatives and morphemes could thus become, at the same time, criteria of efficiency to reduce each grammatical category further to a component of action.

By such means the number of characters became far small than the words of a natural language found in a dictionary, and Lodwick endeavored to reduce this list further by deriving his adjectives and adverbs from the verbs.

From the character to love, for example, he derived not only the object of the action (the beloved) but also its mode (lovingly); by adding a declarative sign to the character to cleanse, he asserted that the action of cleansing has been performed upon the object–thereby deriving the adjective clean.

Lodwick realized, however, that many adverbs, prepositions, interjections and conjunctions were simply not amenable to this sort of derivation; he proposed representing these as notes appended to the radicals. He decided to write proper names in natural languages.

He was embarrassed by the problem of “natural kinds” (let us say, names of substances like cat, dog, tree), and resigned himself to the fact that, here, he would have to resort to a separate list. But since this decision put the original idea of a severely limited lexicon in jeopardy, he tried to reduce the list of natural kinds as much as possible, deciding that terms like hand, foot or land could be derived from actions like to handle, to foot or to land.

In other cases, he resorted to etymology, deriving, for instance, king from the archaic radical to kan, claiming that it meant both to know and to have power to act. He pointed out that Latin rex was related to the verb regere, and suggested that both the English king and the German emperor might be designated by a simple K followed by the name of the country.

Where he was not able to find appropriate verbal roots, he tried at least to reduce as many different sounds as possible to a single root. He thus reduced the names for the young of animals–child, calfe, puppy, chikin–to a single root.

Moreover, Lodwick thought that the reduction of many lexical items to a unique radical could also be performed by using analogies (seeing as analogous to knowing), synonymy (to lament as a synonym of to bemoane), opposition (to curse as the opposite of to bless), or similarities in substance (to moisten, to wet, to wash and even to baptize are all reduced to moisture).

All these derivations were to be signaled by special signs. Wilkins had had a similar idea when proposing the method of transcendental particles, but it seems that Lodwick’s procedure was less ambiguous.

Lodwick barely sketched out his project; his system of notation was cumbersome; nevertheless (with a bare list of sixteen radicals–to be, to make, to speake, to drinke, to love, to cleanse, to come, to begin, to create, to light, to shine, to live, to darken, to comprehend, to send and to name), he managed to transcribe the opening of the gospel of St. John (“In the Beginning was the Word, and the Word was with God . . .”).

Beginning was derived of course from to begin, God from to be, Word from to speake, and so on (the idea of all things is derived from to create).

Just as the polygraphers had taken Latin grammar as a universal model, so Lodwick did the same for English–though his English grammatical categories still reflected the Latin model. Nevertheless he succeeded in avoiding certain limits of the Aristotelian classification of substances, because no previous tradition obliged him to order an array of actions according to the rigid hierarchical schema requested by a representation of genera and species.”

Umberto Eco, The Search for the Perfect Language, translated by James Fentress, Blackwell. Oxford, 1995, pp. 260-3.

Eco: A Priori Philosophical Languages

“The advent of a priori philosophical languages entails a change in paradigm. For the authors we have considered up to now, the search for a perfect language arose from profound tensions of a religious nature; the authors we are about to consider imagined on the contrary a philosophical language which could eliminate the idola responsible for clouding the minds of men and for keeping them afar from the progress of science.

Not by chance, most of the agitation for a new and universal language arose from Britain. There is more to this than a reflection of the English expansion during this period; there was a specifically religious aspect as well.

Although Latin was still the common language of scholars, to the English mind, it was associated with the Catholic church. Besides, it was also too difficult for English speakers. Charles Hooke complained of “the frequent Sarcasmes of the Foreiners, who deride to see such a disability in Englishmen (otherwise Scholars good enough) to speak in Latine” (cf. Salmon 1972: 56).

In the endeavor for a common speech the English had commercial reasons (they thought indeed that a universal language would facilitate the exchange of goods at the Frankfurt fair) as well as educational reasons, since English spelling in the seventeenth century was more irregular than it is today (see Salmon 1972: 51-69).

This was also a period which witnessed the first experiences in teaching language to deaf-mutes, and Dalgarno conducted a number of experiments in this field. Cave Beck (The Universal Character, 1657) wrote that the invention of a universal language would be of advantage to mankind as it would encourage commerce as well as saving the expense of hiring interpreters.

It is true that he added that it would serve to propagate the Gospel as well, but it seems evident that for him evangelization was really just another aspect of European expansion in the new lands of conquest.

He was obsessed, like other linguistic theorists of the epoch, by the accounts of the gestural languages through which the explorers conducted their first exchanges with the inhabitants of those distant shores.

In his account of his exploits in the New World in 1527, Alvaro Nuñez Cabeza de Vaca had complained of the difficulty involved in dealing with native populations which spoke thousands of different dialects, describing how much recourse to the language of gesture had helped the explorers.

Beck’s work contained a frontispiece which showed a European consigning Beck’s project to a Hindu, an African, and to an American Indian who expresses himself with a gesture of his hand.

There was finally the problem of scientific language itself. New discoveries being made in the physical and natural sciences made the problem of finding an adequate nomenclature more urgent, in order to counteract the symbolic and allegorical vagueness of alchemical terms.

Dalgarno confronted this problem in the section entitled “To the reader” of his 1661 Ars signorum: it was necessary to find a language which reduced redundancies, anomalies, equivocations and ambiguities. He specified that such a language could not fail to encourage contact between peoples as well as help to cure philosophy of sophisms and logomachy.

What had long been considered one of the sacred writ’s greatest strengths–its vagueness and symbolic density–was now viewed as a limitation.”

Umberto Eco, The Search for the Perfect Language, translated by James Fentress, Blackwell. Oxford, 1995, pp. 209-10.

%d bloggers like this: