Abstract
On the title page of a tract called Ars Signorum appearing in 1661, its author, the Scot George Dalgarno, announces that he has accomplished something rather sensational (see figure 3.1). Offering a description of what he calls ‘a universal character and a philosophical language’, he asserts that by its means “speakers of the most diverse languages will in the space of two weeks be able to communicate to each other all their thoughts (in everyday matters), either in writing or in speech, no less intelligibly than in their own mother tongues1”. This invention is not only claimed to be a highly effective and easy-to-learn means of communication, but also to provide an excellent introduction to philosophy and logic: “by this means also the young will be able to imbibe the principles of philosophy and the true practice of logic far more quickly and easily than from the common writings of philosophers”. Thus Dalgarno claims to have achieved at least three widely different goals all at once in constructing a language which is suited for universal communication, is capable of being learned quickly, and furnishes a means for enhancing knowledge and logical skills.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Notes
Quotations from Ars Signorum (in references abbreviated as AS) are from the English translation of the Latin text in Cram and Maat 2001 (in references: CM).
Cram traced these papers in the Gregory collection of manuscripts in Christ Church (MS 162, Christ Church, Oxford). Dalgarno’s previously unpublished papers, together with an edition and translation of Ars Signorum and an edition of the other published books and broadsheets, have been printed in Cram and Maat 2001 (=CM).
Most of the relevant papers are preserved in British Library Add. MS 4377. Funke (1929: 46ff.) has briefly explored these papers. The broadsheet advertisements have been printed in Cram and Maat 2001: 83-135, as well as Dalgarno’s correspondence, pp. 413-430.
Cassirer 1923:69 echoes Couturat’s account, writing: “Wilkins, der dieses System zu ergänzen und zu vervollkommnen sucht, stellt statt der ursprünglichen 17 Hauptbegriffe deren 40 auf.
The number of generic concepts effectively used for artificial words reflecting a conceptual hierarchy in Dalgarno’s lexicon is 11, the number of such concepts in Wilkins’s lexicon is 40, but both authors indicate that logically speaking these concepts are not the highest genera. Cf. 3.4.5 and 4.3.8 below.
Thus Funke 1929: 13, 44-45, Funke 1959: 213-214, Cohen 1954: 57 and Salmon 1972: 29.
Thus Asbach-Schnitker 1984: xxv-xxvi, Large 1985: 31-32, Strasser 1988: 212-217, which depend heavily on Slaughter’s account.
Slaughter 1982: vii. Further, on the basis of Cram (1980) she points out that Dalgarno’s approach is fundamentally different from Wilkins’s (1982: 173). 9 Cf. Padley 1985: 325ff.
Salmon 1996: 923-925 is the first general survey treating Dalgarno’s and Wilkins’s approaches as basically distinct, rather than as two alternative routes towards the same end, as a number of authors suggested after Cram 1980 had shown that the two schemes were different.
Christ Church MS 162 ff.21-84. Further references in these sections are by page number in Cram & Maat 2001 (= CM).
On a broadsheet published in 1657 Dalgarno claimed that the universal character he proposed was unlike “the laborious and (as is reported) near infinitely burdensome Chinese script” (British Library, Add. MS 4377, fl44r., CM 85).
Tables of the Universal Character, British Library Add. MS 4377, ff.145-146, CM 87-105).
A detailed explanation of this method is contained in a letter Dalgarno wrote to Pell, December 1657, British Library Add. MS 4377, ff. 150r.-152r., CM 419-423.
The same stroke attached to the middle part and to the lower part of the basic sign indicate the second and third line respectively. For the fourth to the sixth lines, and for the seventh, different strokes are used.
The second to the sixth words are indicated by the same stroke pointing straight left, downwards on the left, and by the same vertical directions on the right part of the basic sign.
For details of the exact method used, see CM 361.
Possibly this insight developed more gradually. The broadsheet News to the Whole World of 1658 shows that Dalgarno was thinking along ordinary lines at that time, claiming that “This Character shall immediately represent things, and not the sounds of Words, and therefore universal, and equally applicable to all Languages”. He adds that “by the same Rules it shall be made effable” (British Library Add. MS 4377, f.l43r., CM 109).
Dalgarno to Hartlib, 20 April 1657. British Library Add. MS 4377, f.l48r., CM 418.
Pell’s and Lodwick’s comments are preserved in British Library Add. MS 4377 ff. 154-159, MS 932 ff. 13-15; letters from Mercator and others to Hartlib are preserved in the Hartlib collection at Sheffield. See Salmon 1966 and CM 14-16 for discussion. Cram 1989 and Cram & Maat 1996 discuss exchanges with Comenius on Dalgamo’s early scheme.
Funke 1929: 37-43 discusses the issue at length, accusing Wilkins of scholarly selfsatisfaction in failing to give the schoolmaster Dalgarno due credit for his contribution. Salmon (1966: 369-370) speculates that Wilkins might have seen a letter Dalgarno wrote to Hartlib, containing some rather unfavourable remarks about Wilkins (this letter has been printed in CM 426-427). Salmon (1972: 31) even claims that this is “no doubt the reason why Wilkins made no acknowledgements to Dalgarno”. Equally speculative, but more straightforward, is the assumption that Wilkins might have been irked by the fact that he was not mentioned by name in Ars Signorum, which does contain an account, be it a brief one, of the previous history of the project, mentioning Ward and Lodwick by name (AS 78-79). It is not unlikely that Wilkins felt Dalgarno’s failure to record his name in that context to be a blatant omission, which he avenged in the Essay.
John Owen (1616-1683) was vice-chancellor of Oxford from 1652 to 1658.
Cram (1994) has argued that this may be explained by the fact that Ward is a representative of a ‘radical’ position regarding universal language, according to which the construction of a satisfactory new language is not impracticable, as the sceptic position claims, but unnecessary.
Cf. Essay, the epistle to the reader (Sig.b2v.): “for the helping of whom [i.e. ‘another person’, i.e. Dalgarno] in so worthy an undertaking, I did offer to draw up for him, the Tables of Substance, or the species of Natural Bodies, reduced under their several Heads; which I did accordingly perform, much after the same Method, as they are hereafter set down”.
Dalgarno wrote to Hartlib (3 November 1659): “I feare my too much friedome & openes while I conversed with men may be by some not made good use of” CM 426). He further asked Hartlib to keep him informed about any publications by Comenius, Urquhart and Wilkins.
Shumaker 1982: 167-170 and CM 140-143 provide English translations.
According to Dalgarno, í derives from b by aspiration (similarly, f derives from p).
Cf. Aristotle, Categories VII, 7b 35-38: “The perceptible seems to be prior to perception. For the destruction of the perceptible carries perception to destruction” (Ackrill 1963: 21).
Cf. Blundeville 1599: 50; Arnauld & Nicole 1970[1662]: 216 (II, XVI).
This essential point has been overlooked by Slaughter (1982: 144), who claims that while Wilkins chose the second method, Dalgarno used the first.
The popularity of dichotomous divisions derived mainly from Ramus (Pierre dela Ramée 1515-1572) whom Dalgarno mentions favourably (AS 31, 46). That division by trichotomy is often equivalent and sometimes preferable to a dichotomous one is pointed out in the logic of Port Royal, whose authors argue that the Ramist zeal for dichotomies is pointless (Arnauld & Nicole 1970[1662]: 214 (II, XV)).
This argument is not convincing, if a species is thought of as a collection of individual things. A set may well be characterized by complementation, i.e., by telling which things are not in it. In using ordinary language, we often make convenient use of expressions characterizing’ species’ by means of ‘negative differences’, such as ‘the rest’, ‘the others’ etcetera. Dalgarno’s point might have been inspired by some vague remnant of essentialism, according to which membership of a species entails having characteristics that are intrinsically connected to that species. On this view, a proper description of a species should mention these characteristics rather than simply stating that members of the species do not belong to some other species. Dalgarno further notes that the use of negative words for positive differences reflects a defect in the vocabulary: thus Latin par, impar is defective, as compared to English even, odd (AS 30). Elsewhere, Dalgarno rejects essentialism (AS 37, 44; CM 368; cf. 3.3.3).
However, the broadsheet containing the lexicon grammatico-philosophicum lists in addition to the category’ spirit’, three categories of obscure status that seem to belong to’ spirit’. These three categories are the ones designated by diphthongs mentioned earlier (cf. 3.4.2). The first of these is called’ spiritual concrete’, which seems to be simply a double of’ spirit’. The other two categories,’ soul’ and ‘angel’, are probably intended to fall under’ spiritual concrete’.
Leibniz later argued against Locke, who made a similar point (Locke Essay, III, VI, §41), that there are good reasons for excluding artefacts from the predica-mental tables: first, these tables only serve to give a general survey of our ideas; secondly, artificial substances do not have the same real unity that natural substances have (NE III, VI, §42).
This remark contrasts with the requirement Dalgarno elsewhere says a proper predicamental series should meet, namely that it has a single highest genus. Probably, by ‘highest genera’ he means those that are highest in the scheme, disregarding the mutual order among these. In fact, as will be explained below, the logical order between the genera that are highest in the scheme is not reflected in their names.
in this case, the number of simple sounds (7) is equal to the number of genera. Two other genera of accidents, namely Peis and Keis, contain more lower genera (9 and 8 respectively). This problem is solved by naming the extra genera’ spas’ and’ spηs’, and’ skas’.
In fact Dalgarno calls them species at AS 52. But in general, and especially in the tables of concrete things, many of the items listed will be considered as genera in most contexts. It was precisely Dalgarno’s intention to construct radical words only for genera; names of the species should be composed out of generic names and words signifying a conspicuous feature of the species. Cf. 3.3.3 above and 3.4.7 below.
in the tables of accidents, none of the lower genera contains more than 9 species, if ‘opposites’ (cf. immediately below) are disregarded, so that no special provisions for excess numbers of things are required.
About a quarter of the places on the table are occupied by more than one radical word. In the table listing the ‘physical concretes’ opposites are absent, with the exception of the lower genera containing parts, i.e. parts of figures, parts of plants, of animals and of artefacts.
I counted 23 instances.
For instance, the genus feim — ‘food and clothing’, rather oddly, comprises two genera: fleim — ‘food’ and fleim — ‘clothing’. Elsewhere there is a list of ‘equipment’ containing species like fam — ‘waterpipe’ and fom — ‘needle’, which clearly are neither food nor clothing. Yet their form indicates that they are subsumed under feim. (For in the tables of concretes, species are designated by the vowel, and intermediate genera by the final consonant). More seriously, the names of the higher genera meis, neis and feis recur as names of lower genera: meis means either ‘mathematical concrete’ or ‘head’; neis either ‘physical concrete’ or ‘mouth’; feis either ‘artificial concrete’ or ‘trunk’.
If’ spiritual concrete’ is regarded as a double of’ spirit’, the number will be 20.
Thus Gerhardt 1890: 7; Couturat & Leau 1903: 15; Cassirer 1923: 69; Cohen 1954: 57; DeMott 1955: 1076; Rossi 1960: 226; Salmon 1972: 29; Slaughter 1982: 145; Asbach-Schnitker 1984: xxv; Large 1985: 31; Padley 1985: 362; Pombo 1987: 76; Eco 1995: 231, to name a few. Funke (1929: 10) notes the discrepancy between the summary table and the tables of the lexicon, but counts, wrongly, 23 genera in the latter (including the word for God, which is not a genus — cf. AS 54 — and the lower genus skas). Elliott (1957: 4) follows Funke in this.
Within parentheses, for strictly speaking, Dalgarno’s table merely says tactus — ‘touch’.
Couturat & Leau (1903: 18) have likewise treated the numerical method exemplified by ‘Íηka — 1.elephant’ etc. as though this was Dalgarno’s, and used the example as an argument showing how difficult it is to learn his language. Thus various commentators have purported Dalgarno’s language to have certain disadvantages which it in fact does not have for the v
in the autobiographical treatise Dalgarno enlarges upon the difficulty of learning numerical orderings. He adduces the testimony of Ward, who said that if someone was to mention one of the ten commandments, he “could not readily tell its numerical order” (CM 369-370).
It is not quite clear what the final s means. In the tables of radical words the following can be found: nνp — aquatic bird; sνm — very much (valde); spi — wing (ala)
For this reason, Dalgarno’s compound words are never ‘taxonomically correct’ as Eco (1995: 234) says they are. On the contrary, since these words combine elements belonging to different, mutually exclusive categories, they destroy the taxonomy.
AS 41. Both of these expressions are cryptic. The best I can make of the second one, nηksofprηk, is ‘whole-footed beast without rising’. The first one, nηkbeisap, provides a good illustration both of the awkwardness of the truncated words denoting parts (cf. 3.4.5), and of the creativity displayed by commentators in trying to make sense of Dalgarno’s artificial words. The few modern commentators who have examined these words in detail have taken the element ap to be the truncated radical word meaning ‘tignum’, which is enumerated under ‘part of a building’ and hence doubtless means ‘beam’. Shumaker 1982: 150 translates ‘tignum’ as ‘trunk’ in an obvious attempt to render the use of ap a sensible procedure. With some reservation, I followed him in this (Maat 1995: 164). Eco (1995: 235) similarly assumes that ap is an ‘architectural metaphor for the proboscis’. Apart from being far-fetched, this interpretation does not make sense of the element beis — ‘mathematical accident’. Shumaker even claims that beis ‘must be a mistake’. It seems much more likely therefore that ap is a suffix indicating superlative (cf. 3.5.3). This yields ‘whole-footed beast, mathematical accident, superlative’ for nηkbeisap, which is not the summit of perspicuity either, but probably intended to express largeness.
For details and discussion, see Cram and Maat 1996.
Asbach-Schnitker 1984: xxvi, Large 1985: 32, Strasser 1988: 216, Eco 1995: 234.
This obviously presupposes a name for ‘pear’ which as yet is lacking. 53 Thus the fundamental text called Technη Grammatikη attributed to Dionysius Thrax (c. 100 B.C.) and Priscian’s extremely influential Institutiones Grammaticae (c. A.D. 500). However, Priscian distinguishes the interjection as a separate word class, and omits the article since this is absent in Latin (Robins 1990: 39, 66)
As usual, the distinction derived from Aristotle, and remained a commonplace throughout the medieval period. Priscian c. A.D. 500 wrote: “The parts of speech according to the dialecticians are two, noun and verb, because only these if joined together make a full proposition, the other parts they call’ syncategore-mata’, that is consignificants” (Institutiones Grammaticae II, 15, quoted by De Rijk 1962: 22 (Vol.1)).
Modistic grammatical theory used an extremely complicated technical vocabulary to cover a number of sophisticated distinctions which it would be digressive to take into account in the present context. Cf. Robins 1990: 83-100 for a brief account, Bursill-Hall 1972 for a more extensive treatment.
Michael (1970: 241-246) to my knowledge is the only modern commentator discussing Dalgarno’s views on grammar at some length. Shumaker 1982: 154-162 provides a useful synopsis of the major characteristics of the grammar of Dalgarno’s language, without adding much relevant commentary on theoretical aspects. Padley 1976 and 1985, though exclusively concerned with grammatical theory, confines his treatment of Dalgarno almost completely to the predicamental foundation of the lexicon (1976: 192-193; 1985: 361-363). However, Dalgarno’s theory of the verb is briefly mentioned (1976: 253; 1985: 306, 375).
Michael accordingly mistranslates the phrase ‘omnis Notio praedicamentalis’ (every predicamental notion) as ‘the idea expressed by each of the predicaments’.
The clause “nullam esse Negationem in terminis simplicibus” (AS 65) is mistranslated by Michael as “negation is never expressed by means of simple terms” which may explain, or be explained by Michael’s erroneous interpretation of it. In general, Michael’s misunderstanding of Dalgarno seems to be due to a failure to take the logical background of most of Dalgarno’s arguments into account. That such an eminent history of grammatical theory as Michael’s book could be written without much regard of the logical tradition might be significant for the relative independence of logic and grammar in the period covered. However, it is understandable that theories such as Dalgarno’s, which are expressly aimed at merging logic with grammar, tend to be misrepresented when viewed solely from the perspective of grammatical theory.
In the autobiographical treatise, Dalgarno notes that he later changed his opinion, and that three, rather than two judicative acts are designated by the copula, namely the ‘dubitative’ act in addition to the affirmative and negative ones (CM 374-376).
It is at present a matter of consensus among historians of linguistics that Chomsky’s claims concerning the Cartesian origin of the distinction between surface and deep structure are incorrect (Chomsky 1966). Cf. Salmon 1969 (= 1988: 63-85), Aarsleff 1970, Keith Percival 1972, Padley 1976: 103-104. As in other instances, the linguistic community has learned from Chomsky’s mistakes, as appears from the ensuing literature on previously forgotten grammarians.
On equivalence, cf. e.g. La Logique ou l’Art de Pensex, II, ch. 4: “Je ne dis den de la reduction des propositions opposees en un meme sens, parce que cela est tout-à-fait inutile” (Arnauld & Nicole 1970[1662]: 162. And Blundeville (1599: 64): “The schoolmen doe give divers rules touching the equivalencie of speeches: but such as in mine opinion are neither necessarie nor profitable”. Of conversion however, La Logique ou l’Art de Penser, II, ch. 17 says that “de-là dépendent les fondements de toute l’argumentation” (Arnauld & Nicole 1970[1662]: 219).
Michael is clearly making a similar point, though in somewhat obscure terms, when he says: “His [Dalgarno’s] contention is useless because in so far as a relation is being considered only as expressing a name it is prevented from fulfilling its only task — to relate. Putting it crudely, if the copula is regarded only as a form of the name junction it cannot join” (1970: 245-246).
He does not however reject the distinction between composition and derivation altogether, saying that compounds formed from preposed mutilate particles like ‘re-’ and ‘con-’ in Latin are more correctly called derived than compounded (AS 71), without explaining why this should be so.
This instance shows that Dalgarno sometimes uses interposition of h rather than r to indicate opposition, namely in cases where this is more euphonical. Cf. 3.4.5 above.
Although this example and others may create the impression that what Dalgarno presents as the logical structure of speech is in fact the structure of English as opposed to Latin, there are other instances in which he contrasts the structure of English expressions with logical structure.
Dalgarno also notes that if ‘to be’ is inflected for tense, and hence the proposition is not eternal, or to put it more correctly according to Dalgarno, the proposition is not a proposition of necessary connection, ‘to be’ is to be rendered by dan ‘time’ plus a suffix indicating tense, e.g. Petrus danesa ‘Peter was’ (AS 76-77).
This particle is derived from a radical word meaning ‘pertain’ or ‘belong’. Shumaker 1982: 155 draws attention to a parallel with Pidgin English having ‘hand-belong-master’ for ‘master’s hand’.
This has escaped Shumaker’s usually alert attention, for which reason he finds the paragraph on impersonal verbs obscure (1982:156).
The existence of impersonal verbs had led Humanist grammarians to note a discrepancy between surface and deep structure. Accordingly, Linacre (c. 1460-1524) and Sanctius (1554-1628) assumed a ‘nominative of related signification’ to account for the equivalence: “pluit = pluvia pluit” (it rains = the rain rains). Padley 1976: 105.
Dalgarno’s use of the term ‘integral’ in the autobiographical treatise may well be derived from Wilkins’s terminology. The point here is merely to underline that the traditional distinction expressed in these terms was taken for granted by both Dalgarno and Wilkins. Leibniz made a similar distinction (cf. 5.5.3).
The first word is the Latin particle, the italicized words are the equivalents in the philosophical language, the words in parentheses are the Latin words defining the meaning of the radical words in the lexicon which function as particles here; in square brackets the English translation of Latin words is given.
Since shνb also means ‘post’ (after), and following the syntactic rule that the noun precedes the adjective.
it would therefore be illusory to look for a connection, in the style of Boole, between logical conjunction and logical product here. Neither does Dalgarno associate disjunction with addition, in view of his translation of ‘or’ (aut and vel). Likewise, Shumaker’s suggestion that “the addition of something is construed as multiplication, if only by two” is far-fetched though amusing (1982: 158).
Since unusquisque is not in the list of 82 particles, this remark may be taken to show that Dalgarno did not think that a special convention for particles like these was necessary.
“in a late treatise of near affinity with this [i.e. Didascalocophus, 1680] I promised to give some reasons why an Universal Character so much desired by the learned of former ages when discovered and brought to light and to greater perfection too than ever was thought possible, yet should meet with so cold entertainment by the learned of the present age. Its first appearance upon the stage from diverse hands and in diverse dresses made learned men suspect it for a Proteus or some greater Monster” (CM 372).
Couturat & Leau 1903: 18; Funke 1929: 12-13,14; Strasser 1988: 217.
To no avail, in the view of Funke (1929: 13), who states that this method adds to the difficulty.
Funke’s other major objection is that Dalgarno’s grammar consists almost completely of elements borrowed from existing languages, which is, Funke contends, because Dalgarno had no idea that a philosophical language should designate elements and structures of consciousness (1929: 14). This objection blatantly overlooks the principal objective of Dalgarno’s grammar.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2004 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Maat, J. (2004). Dalgarno: The Art of Signs. In: Philosophical Languages in the Seventeenth Century: Dalgarno, Wilkins, Leibniz. The New Synthese Historical Library, vol 54. Springer, Dordrecht. https://doi.org/10.1007/978-94-007-1036-8_3
Download citation
DOI: https://doi.org/10.1007/978-94-007-1036-8_3
Publisher Name: Springer, Dordrecht
Print ISBN: 978-94-010-3771-6
Online ISBN: 978-94-007-1036-8
eBook Packages: Springer Book Archive