Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

Given that the vast majority of artificial languages (henceforth ALs) have seen very little, if any, use, one might have serious doubts about whether there could be much of a pragmatic study of them. Indeed, there has been very little pragmatic research on ALs, which have generally received little attention from theoretical linguists (who might feel that they are not serious objects of study, a matter not helped by the perhaps biased work of some Esperantists and by the existence of “languages” such as Klingon), but some linguists may believe that there would be even weaker grounds for pragmatic analysis of ALs than for e.g. morphological or syntactic analysis, since most ALs have not been used much, if at all.Footnote 1

However, many works presenting an AL do provide (sometimes extensive) textual material, which one can analyze pragmatically (as well as syntactically). Since such texts were largely intended as instructional material, one might say that they do not represent a typical use of language, but they do represent a kind of use (and one meant to mirror more typical uses). It could be claimed that even example sentences and reading exercises constitute a certain kind of language use. One of the few works treating artificial language pragmatics is Huang (2002), but, in spite of the fact that “Artificial Language” is contained in its title, the discussion of ALs makes up only a small part of it. More space is given to ALs by Traunmüller (1991/1996); however, some of the phenomena which he deals with (e.g. number marking in NPs) are not those usually dealt with by pragmatics.

If we now look at particular views on pragmatics and ALs, Galdia (2009) would probably deny the possibility of a serious pragmatic study of ALs, as he says (p. 334):

No natural language has been meaningfully characterized without the pragmatic dimension nor can it be properly spoken without it. The lack of knowledge about this dimension makes a full characterization (and also the full command) of ancient languages like Latin or Greek impossible. This theoretical problem is even more manifest in the case of artificial languages such as Esperanto which cannot be fully determined in terms of pragmatics simply because they are artificial. An artificial pragmatics, in turn, would not make much sense.

One may compare this opinion with the remarks of van Cranenburgh et al. (2010) about Esperanto: “Although it was designed as an easy-to-learn language, with regular and transparent syntax and morphology, its semantic and pragmatic components have evolved naturally” (p. 2); “Esperanto has a regular and transparent morphology while featuring rich semantics and pragmatics” (p. 7). Similarly Dellert (2008: 2) states:

since the language has developed into a full replacement for natural languages in all situations, all the aspects of semantics and pragmatics that NLP [natural language processing] wants to address are present in Esperanto as much as in any natural language

About ALs more generally Gobbo (2008: 39) says:

From the point of view of theoretical linguistics, planned languages are fully human languages, being non-natural without necessarily being unnatural, since they are acquired or acquirable as a normal part of the process of maturation and socialization (Lyons 2006). Consequently, they will be scrutinized by linguistic level, from language core to language use, that is to say phonetics, morphology, syntax, semantics, pragmatics.

One might think that if one were going to pragmatically analyze an AL, Esperanto would clearly be the best choice, since it has seen far more use than any other ALs (and in fact has some first language speakers), and second choices would be ALs which have been used to a relatively high degree (for ALs) such as Ido and Interlingua, while it would be unproductive, if not silly, to try to treat ALs which have been used very little or not at all (i.e. the vast majority of them).

However, although I shall discuss Esperanto and Ido, I take a different position. ALs are commonly classified on the basis of how much material, if any, they take from natural languages. Those which, like Esperanto, are largely or entirely based on one or more natural languages are known as a posteriori ALs, and it is these which have been most successful and popular. Those which represent attempts to build a language from scratch, i.e. for which little or nothing is (consciously or intentionally) taken from natural languages are called a priori ALs. This is a spectrum rather than a strict dichotomy: most a priori languages have borrowed at least a small amount from natural languages, and many languages have substantial amounts of both original and a posteriori material; such languages are referred to as mixed ALs.

In the early days of AL language construction, i.e. in the 17th and 18th centuries, most ALs were designed along a priori lines, though later the mixed and a posteriori methods of language creation became much more popular. Nevertheless even in the 20th century some a priori languages, such as Ro, Suma, and aUI, were constructed. No a priori language has come close to the level of use of Esperanto or Volapük, the most successful mixed language.

However, it might be argued that if one is interested in the pragmatic possibilities of human language, the languages to look at would be the a priori languages, for these would be good testing grounds for putative pragmatic universals. If an AL is meant to be a completely original creation, owing nothing to natural languages, and if even in it such universals hold, then we may have an idea of the limits of pragmatic variation.Footnote 2

Another general area where we may be able to learn something from ALs, of whatever sort, is laymen’s ideas about pragmatics. The vast majority of AL designers or describers are not well-versed in linguistics, and what they say (or perhaps more importantly, do not say) about pragmatic features of the language can tell us what the average person thinks and knows about pragmatics.

It may be significant for our purposes that ALs are usually second languages, and so there may be pragmatic (and other) influence from a first language. Given this, one might ask which texts in an AL would be suitable for pragmatic analysis; that is, presumably only texts created by those with a certain degree of proficiency would be of interest (unless we were specifically interested in early stages of the learning process)—what is that level and how can we determine it, or can we? (The same issues come up if we are analyzing texts at some other level, e.g. for syntax, but there may be a need for a larger amount of textual material if one is looking at pragmatics) Most ALs do not have official (or any) courses which could give some certification or evidence of having reached some level of skill in the language. It would appear reasonable to use texts written by the AL designer (and sometimes the only texts in an AL are by its designer), although sometimes even they make errors in their languages. It would also appear reasonable to use texts in material which in some way is (explicitly or implicitly) approved of by the designer (e.g. in webpages which he gives links for in his website) or by some official organization or webpage for the AL.

I shall deal with the following areas: politeness, different styles and levels of language (including formal language), non-descriptive meaning, illocutionary force, and non-literal language. I shall thus not be discussing presuppositions, because I have found neither many relevant examples nor any significant discussion of them by AL designers.

2 Politeness

Politeness may be the most obvious pragmatic feature of language and the one which most often comes up in presentations of ALs (compare e.g. conventional implicature, which most AL designers, like most speakers of any language, are probably not consciously aware of).

2.1 Pronouns

ALs, like natural languages, differ in whether they have polite pronouns. Several AL designers explicitly state that their language lacks them. For example, Russell (1966: 5) says of Suma, “There is no polite form of the pronoun ‘you’”.

Bollack, the designer of the Blue Language, believes that an AL, or at least his AL, must have 2nd person polite pronouns (the Blue Language has not only a singular, but also a plural, polite form); he states (1900: 19), “This creation is absolutely necessary, as the «civilization», when to address somebody, imposes on us a familiar and a respectful form”.

Not all presentations of ALs with polite pronouns give instructions on how or when to use them (although the same is true of many pedagogical grammars of natural languages). For example, Talmey (1925: 7) gives vu as the “singular, polite” 2nd person pronoun of Arulo (cf. tu ‘thou, thee, you’ and vi ‘you (plural)’), but says nothing about situations in which it should appear. A lack of instructions could be seen as unfortunate, since it is not obvious in which situations they are appropriate, especially for native speakers of languages such as English which lack polite pronouns. Since the use of polite pronouns differs from language to language, even native speakers of languages with such pronouns may be uncertain about this feature of an AL.

De Beaufront (1925/2005) is a source which does give instructions about them; he says (p. 25) concerning Ido (which has three 2nd person pronouns, two singular ones, tu and vu, and the plural form vi), “Por la duesma persono singulara existas anke formo familiara: tu, quan on darfas uzar nur ad amiki tre intima, a frati o parenti kun qui on uzas, en sua linguo matrala, formo familiara korespondanta” (‘For the second person singular there also exists a familiar form, tu, which one may use only with very intimate friends, with siblings or relatives with whom one uses, in his mother language, a corresponding familiar form’).

ApGawain et al. (2008: 37) have more complete instructions on the use of tu in Ido:

  1. 1.

    “tu” refers to one person only. It shows affection towards the person addressed, and is therefore only to be used in special circumstances: (a) within the family, (b) between close friends, (c) when addressing small children, (d) perhaps when addressing an animal or pet.

  2. 2.

    “vu” also refers to one person only. It is the usual word for “you”.

  3. 3.

    “vi” refers to more than one person, and is the plural of both “tu” and “vu”.

Concerning Sambahsa Simon (2010: sec. 3) says:

“Yu” can apply to a group of persons or to a single person as a sign of formal respect. “Tu” (cf. archaic English “thou”) is used only to address close relatives or friends and children, as, for example, in French, Russian or Farsi.

Parke (2008: 9) gives a rather detailed description of the use of the familiar and polite pronouns of Frenkisch:

Unlike in English, there are both plural and singular forms for the second person pronoun, ji and dou. Ji should always be used when addressing more than one person. When addressing one person in a formal context, ji should also be used. Dou is a more familiar, intimate and informal pronoun than ji. Dou should be used mostly for addressing friends, lovers, family members and young children. Its use may be extended to colleagues and fellow students but then care should be taken to use it only informal situations. Used in the wrong situation, dou can be seen by some to be rude or excessively intimate—the verbal equivalent of invading someone’s personal space.

Esata has both a 2nd person singular pronoun, yu, and a 2nd person plural and formal pronoun, yi, but it also has a sort of honorific affix for pronouns; Bothi (2006: sec. 6) says, “Deference and respect can be indicated by prefixing (stronger) or appending the pronoun modifier ji: e.g. yuji—your honor, jiyu—your excellency”. It is interesting that the level of respect it marks differs depending on its position.

In Latinulus various pronouns can replace other pronouns; Martellotta (1919: 103) gives the following instructions on pronoun usage:

  • Il pronome di prima persona plurale viene adoperato in vece del singolare dai sovrani, dai magistrati e dagli scrittori in senso maiestativo.

  • Il pronome tu si usa parlando fra colleghi, o con inferiori di dignità, o ad animali, o ad esseri sacri.

  • Il pronome di seconda persona plurale si usa sovente invece del singulare quando si parla o si scrive ad una sola persona.

  • Il pronome di terza persona singolare e plurale leis e leise si usa invece del pronome di seconda quando si parla o si scrive ad una persona di riguardo.

  • (‘The pronoun of the first person plural is used instead of the singular by sovereigns, by magistrates, and by writers in a majestic sense.

  • The pronoun tu [‘you’] is used when speaking among colleagues, or with those lower in rank, or to animals, or to sacred beings.

  • The pronoun of the second person plural is often used instead of the singular when one is speaking or writing to a single person.

  • The pronouns of the third person singular and plural leis [‘she’] and leise [‘they’ (feminine)] are used instead of the pronouns of the second person when one is speaking or writing to a person of regard.’)

Esperanto has a 2nd person pronoun in addition to vi, but it is very rare and its function is not clear; Wennergren (2005: 102) states:

Ci estas unu-nombra alparola pronomo… Ci kaj cia ekzistas nur teorie, kaj estas preskaŭ neniam praktike uzataj. Eblus imagi ci kiel pure unu-nombran vi, aŭ kiel intiman familiaran (unu-nombran) vi, aŭ eĉ kiel insultan (unu-nombran) vi. Sed estas fakte tute neeble diri, kian nuancon ĝi montras, ĉar ĝi apenaŭ estas uzata

(‘Ci is a singular addressing pronoun… Ci and cia exist only in theory and are almost never used in practice. One could imagine ci as purely a singular of vi, or as an intimate familiar (singular) vi, or even as an insulting (singular) vi. But it is in fact completely impossible to say what nuance it indicates, because it is hardly ever used’)

This raises the important point that one cannot know all of the details of a word which does not occur, or, to see it another way, it is only through use that a word acquires all of its meaning. (This recalls the argument (which I do not fully agree with) that there cannot be a pragmatics of an artificial language, or at least of most artificial languages, as they saw little or no use.)

However, Kellerman (1910: 15) gives a different impression of ci:

There is another pronoun [besides vi] ci, thou, for the second person singular, used in solemn style, as in the Bible, in poetry, and also for intimate or familiar address when desired, like German du, French tu, etc.

Gledhill (2000:103) says the following about this word:

The pronoun ci was originally devised by Zamenhof to represent the archaic informal or Shakespearean ‘thou’ in translations and is used by some Esperantists as a familiar ‘you’ because of the influence of European languages (French tu, German du, Russian ti etc.). Zamenhof discouraged ci with the justification that different languages had different conventions for its use, and there may also have been an ideological motive to establish parity for all language speakers. There are no examples of ci in the corpus,Footnote 3 although its use is attested among some reform-minded Esperantists. Conversely, a polite form Vi (i.e. capitalized vi) was widely used as a polite or plural second person form in early writing, and there are a number of instances in the literary sections of our corpus, including the early writings of Zamenhof.

There are some interesting points here. The fact that some natural European languages have played a role the occurrence of ci indicates that at least in this area of pragmatics ALs can be affected by natural languages. (One might have assumed that this could happen, but one also might want evidence for it, as we now do.) Also intriguing is the connection between ideology (in this case egalitarianism) and a view about pragmatic feature, the familiar/polite distinction in pronouns. It would appear that there is some sort of impulse, at least on the part of some language speakers, to have a polite pronoun, since there has been both the aforementioned effect from European languages with the respect to ci and the creation of the other polite pronoun Vi.

The situation for Volapük’s polite pronoun ons is somewhat similar to that of Esperanto ci; Linderfelt (1888: 14) states:

The English “you” having usurped the functions of both “thou” and “ye”, observe that in Volapük ons represents a polite “you”, whether addressed to one or more persons, ol the “you” of familiar intercourse to one person and ols to two or more persons. The ons is, however, of doubtful value and might as well be dropped, the English usage having amply demonstrated that one form of address may be employed to everybody, without giving offense to anyone, though keeping a sharp distinction between the singular ol and and the plural ols is indispensable to clearness. At the Volapük conference in Münich, it was decided to discourage the use of the form in ons.

There are thus different views about whether ALs should have polite pronouns.

2.2 Forms of Address

As with polite pronouns, sources on ALs sometimes do not give instructions on when to use forms of address. Again de Beaufront, on Ido (1925/2005), does give instructions; on p. 184 he deals with salutations in letters. He begins by noting the need for establishing standard practices:

Ta formuli esas afero di nacionala kustumo e stilo, e la simpla traduko di tala nacionala formuli genitus ne nur senfina diverseso, ma frazi stranja, nekomprenebla o miskomprenebla. Semblis do necesa fixigar, per konvenciono, to quo devas konsideresar kom polita formuli.

(‘These formulae are a matter of national custom and style, and the simple translation of such national formulae would generate not only endless diversity, but sentences which were strange, incomprehensible, or liable to be misunderstood. It therefore seemed necessary to fix, by convention, that which should be considered as polite formuli.’)

He then gives details:

Ye la komenco di letro, ni uzez nur Sioro, e se la korespondanto havas ula titulo, funciono o profesiono, qua konsideresas en nia korespondado, ni skribez: Sioro Prezidero, Sioro Profesoro, e. c. Por iti, qui havas funciono o situeso, por qua la simpla Sioro ne semblas suficanta, ni havas la vorto Sinioro: Sinioro Episkopo, Sinioro Ministro.

Ye la fino di letro, ni generale uzez: Kun sincera saluto. Por siniori e la personi, quin ni qualifikas «sinioro» ni dicez: Kun respektoza saluto.

(‘At the beginning of a letter, let us use only Sioro [‘Sir/Madam’], and if the correspondent has any title, office, or profession which is considered in our correspondence, let us write: Sioro Prezidero [‘President’], Sioro Profesoro [‘Professor’], etc. For those who have an office or situation for which the simple Sioro does not seem sufficient, we have the word Sinioro [‘Lord, Highness’]: Sinioro Episkopo [‘Bishop’], Sinioro Ministro [‘Minister’].

At the end of a letter, let us generally use Kun sincera saluto [‘With sincere greeting’]. For nobles and people whom we term “noble” let us say: Kun respektoza saluto [‘With respectful greeting]’.)

Finally he states that these instructions do not apply to informal situation: “Komprenende ta reguli ne koncernas la korespondado kun amiki, kamaradi, parenti, qua admisas tre granda diverseso en ta formuli.” (‘Of course these rules do not concern correspondence with friends, comrades, relatives, which allows very great diversity in these formulae.’) (On pp. 16–18 he treats forms of address more generally.)

Anderson (n.d.: 120) also discusses how to open and close letters:

The practice of using “Dear (Sir/Madam)” as an introductory term in general correspondence is as unfitting in its expression of quasi-affection as “Sir/Madam” is in its curtness. Some form of the word “Respect” is seemingly appropriate here. The introductory “(Dvm/Made) li hail”: “(Sir/Madam) in respect” can be recommended; and this also in public address. When a Christian name is employed, then “– li zan”,Footnote 4 or other more or less endearing expression may be deemed appropriate. As a completory, and with a like discretion: “Ua qu (li) yer/yrv/aye/sain bvn si”: “To you (with/in) truth/earnestness/love/best wishes” have a simple dignity.

Another author who deals with this matter is Martines d’Antoñana, the designer of Neoispano; he writes (in Neoispano) (p. 91):

O terminologia do korespondensia, elemento tan importante en o komunikasión sosial i komersial, se modernisa en neoispano adoptando expresiones piu práktiko ao prinsipo i fin d letras. O expresiones, aktualmente en uso d “Muy señoro mío” i “Su seguro servidor”, es no solomente antikuado, sino beromente ridíkulo, porke no tiene ningún signifikado. En korespondensia komersial, Señor: ao prinsipio, i, Salute. ao fin, es piu apropriado.

Letra ao familia i relatibos: Kerido. Pa amikos: Estimado. Extraños: Señor. Ao fin d letra: Afektuosamente, Kordialmente, etc. Extraños: Salute.

(‘The terminology of correspondence, an element so important in social and commercial communication, is modernized in neoispanso, adopting more practical expressions at the beginning and end of letters. The expressions currently in use, “Muy señoro mío” [lit. ‘my very (much) sir/lord’] and “Su seguro servidor” [‘Your certain servant’] are not only antiquated but truly ridiculous, because they do not have any meaning. In commercial correspondence Señor [‘Sir’], at the beginning and Salute [‘Greeting’ (?)] at the end is more appropriate.

Letter to family and relatives: Kerido [‘Dear’]. For friends: Estimado [‘Esteemed’]. Strangers: Señor. At the end of letter: Afektuosamente [‘Affectionately’], Kordialmente [‘Cordially’], etc. Strangers: Salute.’)

It is perhaps worthy of note that these authors give such importance to a relatively minor area of pragmatics (while saying nothing about some other issues).

ApGawain et al. (2008: 76) say the following about Ido forms of address:

  1. 1.

    Sioro (Sro): Mr/Mrs/Miss/Master/Sir/Madam. This can be used to address either a man or a woman, married or single, irrespective of age. For example in a business letter:

Estimata Sioro—Dear Sir/Madam

  1. 2.

    Siorulo (S-ulo): Mr/Master/Sir. In practice this is not often used, Sioro being sufficient.

  2. 3.

    Siorino (S-ino): Mrs/Miss/Madam

In practice Sioro is not often used for women, Siorino being prefered. This is partly as a compliment to the female gender and partly to help distinguish between different members of the same family: Sro e S-ino Smith—Mr. and Mrs. Smith

Any woman has the right to be addressed as Sioro, should she so desire it. Note that Siorino can refer to both married and single women.

  1. 4.

    Damzelo (Dzlo): Miss/young lady:

Should circumstances require it, an unmarried lady may be addressed as damzelo: Damzelo Jones, Yen S-ino e Dzlo Smith—Here are Mrs. and Miss Smith. La damzelo (qua esas) ibe—The young lady over there.

It is interesting that they speak of the “right” to be addressed in a certain way.

2.3 Honorifics

Few ALs have honorifics. This is not surprising since a large proportion of ALs are based on one or more Western European languages, which lack honorifics. However, there is an honorific in Sona. Searight (1935: 33–4) says the following about it, and about the need for it:

Sona has a special form of Article called the Honorific, borrowed from J. [= Japanese]. This is the vowel o; … It is used before names, forms of address, and verbs as an expression of politeness. Thus we have:- o ra ‘(honorable) man’, ‘gentleman’, o hara ‘sir’, o tu jiko ‘your (honour’s) children’, o toru ‘please pass’, o min ‘(please) come in!’ We meet with so many ways of address in national languages, ranging from the flowery honorifics of the East to the laconic ‘Say bo’ of the new West, that we must have some such mechanism to satisfy all tastes.

The Italian has no less than three forms of address—tu, voi, lei [sic]; the haughty Pathan but one—‘ty’—whether to prince or pauper. In English we write Mr. to our tailor, but Esq. to our friends—yet neither word can we use in address. We have no happy way of calling the waitress. We hover between a furtive ‘Miss’ or a self conscious cough. In Sona the simple little vowel o solves all our difficulties. The word hara, both in writing and address, means ‘Mister’, whether tailor or friend, while o hara covers all the complications of Sir, Dear Sir, Respected Sir, Your Honour, and so on. Likewise tu ‘you’ is exalted to o tu in polite address, and solves the problem of ‘Yours Truly’, ‘Yours Faithfully’, ‘Your Obedient Servant’, and all the rest.

Ardano has the same honorific marker, but apparently it is only used with questions; Elhassi (2008: Lesson 10) says:

- Respectful sentences:

We add (O) as a prefix before [t]he interrogative

The idea is taken from Japanese.

Hal → Ohal

Hal ti posna aiutije min? = Can you help me ?

Ohal ti posna aiutije min ? = Coulld you help me?

Canjan → Ocanjan

Canjan ti farna? = How are you?

Ocanjan ti farna? = How do you do?

2.4 Imperatives

Some ALs have more than one type of imperative, the difference between/among them involving politeness or something like it. This is true of Volapük; Sprague (1888: 26) says:

The ending of the simple imperativ [sic] is öd, following the person-ending. […] There are two modifications of the imperativ, the courteous or softened form in -ös and the harsh form in -öz; called by some grammarians the optativ [sic] and the jussiv. The former expresses a request and the latter a positiv [sic] command.

In the exercise which follows this (p. 27) and the key to it (p. 118), Sprague usually uses please as the English equivalent of -ös and an exclamation point as the equivalent of -öz (sentences containing forms in -öd and -ös end with a period rather than an exclamation point), e.g.:

(1a) Kömolsös al visitön obis in dom obas nulik.

‘Please come to visit us in our new house.’ (p. 27, tr. p. 118)

(bb) Gololöz se dom!

‘Go out of the house!’ (p. 27, tr. p. 118)

Couturat and Leau (1903: 377) say about Kosmos:

Pour l’impératif, on emploiera le subjonctif présent (forme polie): amösi, aime; amösis, aimez; pour un impératif plus bref et plus pressant, on emploiera le radical verbal en -o (avec -s au pluriel): curro, cours; venios, venez.

(‘For the imperative, one will use the present subjunctive ([as] the polite form) amösi ‘love’; amösis ‘love’ [pl]; for a brief and more insistent imperative on will use the verbal root in -o (with -s in the plural): curro ‘run’; venios ‘come’ [pl.].’)

Eichhorn’s Weltsprache also has two imperative forms; Couturat and Leau (ibid.:300) state, “L’impératif se forme en ajoutant le pronom personnel à l’infinitif, et en intercalant un ü ou un i, suivant que le sens est plus ou moins impérieux” (‘The imperative is formed by adding the personal pronoun to the infinitive, and inserting an ü or an i, according to whether the meaning is more or less imperious’).

2.5 Word Order

In Ande word order can have a polite function; Anderson (n.d.:97) says:

Standard order is: Subject, Predicate, Object, Indirect Object … However, as sentential precedence may convey a suggestion of respect, or indulgence, it may be desirable on occasion to promote the Object. For example: “Me zan qu nu aya sio” : “Darling I love you so”; (Lit.—you I love so (much)).Footnote 5

Although of course word order has pragmatic functions in language, I know of no natural (or other artificial) language in which it has this type of function.

3 Formal Language (and Other Levels of Language)

One might be surprised to learn that in some ALs there is, or is supposed to be, more than one level of language: although it is not very common, AL designers sometimes make remarks about words or constructions being appropriate for one kind of language.

Quiles (2009) makes several remarks about what should be, or is, used in a formal style of Modern Indo-European:

In Modern Indo-European, compounds may be written with and without hyphen, as in the different modern Indo-European languages. Nevertheless, the older, not hyphenated version is preferred for formal writings; as, sindhueurōpājóm, and not *sindhu-eurōpājóm [‘Indo-European’] (p. 126)

The plural wejes [‘we’] is often used for the singular \( \bf{eg} \dot{\bar{\bf{o}}}\) [‘I’]; the plural juwes [‘you’] can also be so used for the singular tū. Both situations happen usually in formal contexts. (p. 183)

Formal writings in Modern Indo-European should follow the patterns attested in the oldest inscriptions, i.e. (S)OV, as in Vedic Sanskrit, Ancient Greek, Old Latin and Avestan. (p. 266)

In Modern Indo-European, thus, negation should usually be preverbal, as in modern Romance languages (cf. Fr. n’est, Spa. no es, etc.), but it can be postponed in emphatic contexts, as it is usual in modern Germanic languages (cf. Eng. is not, Ger. ist nicht, etc.), as well as in very formal texts, thus imitating some of the most archaic findings of early PIE dialects. (p. 300)

Stadelmann (1945: 36) indicates that there are several registers in Voldu:

In literary, scientific or commercial language it might be desirable to omit the personal pronoun. In this case the verb is conjugated like that:

Present: Eleh, elez, eles, elek, elec, elet.

Past: Elah, elaz, elas, elak, elac, elat.

Imperat[ive]: Eluh, eluz, elus, eluk, eluc, elut.

Ex: Roma(n) regoy governat. Kings governed Rome.

(Rome was governed by kings) (Latin style).

Da noktes. Is [sic, presumably should be It] grows night. (Poetic style).

Vayuk! Let us go!

Eluz man! Be a man!

Elus kyet! (Hi sol el kyet!) He is supposed to be quiet!

[…]

Sun zaynes. The sun shines. (Poetic style).

Not all of this is completely clear, but the basic idea seems to be that verbs can bear agreement marking in certain kinds of language; generally they do not, thus for example the present tense form of the verb el ‘to be’ is el, no matter what person or number the subject is, and the past tense form is ela. The agreeing forms are used when the subject pronoun is dropped. Given the first and last examples, it appears that they can be used even when there is an overt subject, as regoy ‘kings’ and Sun ‘The sun’ are the subjects of these sentences.Footnote 6

The second example seems to be of the same type; one might think that Da is an error for Dag ‘day’ and is thus the subject, while noktes is a verb form derived from nokt ‘night’, specifically the 3rd person singular present tense form. This example might therefore be more literally translated as ‘The day nights (i.e. becomes night)’.

The remaining examples involve imperatives. 2nd person imperatives do not have to have (and perhaps cannot have) an overt subject, e.g. Kam tu mi! ‘Come to me!’ (ibid.:43), but they also usually do not have any suffixes, i.e. the imperative form is identical to the infinitive. However, from the passage above we see that they can have an ending. One can express 1st person plural imperatives with the auxiliary verb lar ‘let’ and the infinitive of the main verb, e.g. Lar nun vay! ‘Let us go!’ (ibid.:64), and this is perhaps what is done in normal style, but there is the option of having marking 1st person imperatives synthetically (as well as 1st person singular imperatives). The same general situation seems to hold for 3rd person imperatives, as shown by the alternatives given in the passage, Elus kyet! and Hi sol el kyet!

Stadelmann (1945) also makes some remarks about “familiar” language: on p. 24 he says, “In familiar speech the ke of a conjunction can be dropped” and on p. 36 he states, “Contractions can be imitated from English in the following way (Familiar style): Ex: Yu’l n’t. (Yu el not). [‘You are not’] Hi’av n’t. (Hi hav not). [‘He does not have’]”.

Simon (2010: sec. 3.4) says, “Due to its Indo-European heritage, Sambahsa has some verbal forms only encountered in literary usage”. One might think this odd, since various other ALs based (largely) on Indo-European languages do not have such forms, and in any case Simon (the designer of Sambahsa) was under no compulsion to retain such forms. These forms include some alternative person agreement endings:

If these are compatible with the accentuation, verbs can bear these endings in the present indicative:

  1. 1°

    person singular: -mi

  2. 2°

    person singular: -si

  3. 3°

    person singular: -ti

  4. 4°

    person plural: -nti.

The corresponding forms for «ses» [‘to be’] are esmi, essi, esti, sonti.

The other conditions for the use of these forms are that the verb stands [sic] in absolute initial position in the clause, and that this clause contains [sic] no adverb. Those conditions are seldom fulfilled. (ibid.)

That is, Simon has created forms which will rarely occur, which is reminscent of the situation with Esperanto ci. One might wonder why he went to the trouble of doing this. In the same section Simon also brings up two sets of “old forms”, imperatives and infinitives. They cannot literally be old, since Sambahsa itself does not seem to be very old (its presence on the internet dates from 2007), but perhaps they are meant to have the feel of archaic language.

There are forms from other word classes which are also used in particular circumstances; Simon (ibid.:sec. 2.2) states:

In Sambahsa-mundialect, endings with declensions can be added to substantives and adjectives for purposes of euphony or literary purposes (ex: poetry). This system, whose native name is euphonic vocalisation, can only be used if it is compatible with the accentuation patterns. For example: uno smiegdo geront “a frail old man” instead of un smiegd geront. In everyday use, those endings appear only in the words vasyo (all of the, every) and alyo (another).

In Esperanto there are some “unofficial” elements (e.g. the suffix -ator-, which forms words for machines). Although one would not want to interpret their existence to mean that there are two levels of the language, an official and an unofficial one, there are different elements with a different status and the language is not completely under prescriptive control. This of course is the situation with natural languages (in spite of organizations such as the French Academy), and is perhaps to be expected of an artificial language which sees a large amount of use. Such facts could be taken as support for the idea that there can be a pragmatics of an AL, as an AL can be used extensively and is to some extent under the control of its users, i.e. it is not just an abstract and sterile creation.

An interesting question about formal language in ALs is which features are considered to be formal; given that an AL designer has complete control of his language (at least at the beginning), he can label any words, contructions, etc. that he wants as formal (literary), etc. One might wonder what effect his native language (or any other language) could have on this.

4 Conversational Implicature

I have found no explicit discussion of conversational implicatures in instructional materials for ALs, and I have not found many examples of them in AL texts. Here is one example in Usik, from Palanca Gómez (2008: 126), in a pseudo-dialogue in which someone is reporting an accidentFootnote 7:

Bulki bam?

‘Is there any wounded?’

Bel, ank tendi solno bo nuski leabi

‘Well, a woman lying down the floor that she breathes troublesomely’ [sic]

Here the maxim of relvance is involved; the answer does not directly mention anyone who is wounded (and one could have difficulty breathing without being injured). Notice that it is introduced by bel ‘well’, which apparently can be a marker of an utterance involving a conversational implicature, as well can be in English.

5 Non-descriptive Meaning

5.1 Conjunctions

Sources on ALs usually are not explicit about the non-truth conditional meanings of some conjunctions. For example, a word may simply be glossed as ‘but’. However, Wennergren (2005: 304–5) goes into detail about the uses of the Esperanto word sed ‘but’:

Sed ligas frazojn aŭ frazpartojn, kaj montras, ke ili iel kontrastas inter si: … Sed povas ankaŭ enkonduki ion, kio malebligas aŭ malhelpas la antaŭan aferon: … Sed povas enkonduki ion, kio estas surpriza, se oni konsideras la antaŭan aferon: … Se la antaŭaj vortoj esprimas neadon de io, sed povas enkonduki tion, kio validas anstataŭe: … Iafoje sed montras kontraston al io subkomprenata

(‘Sed connects sentences and parts of sentences, and shows that they contrast among themselves: … Sed can also introduce something which makes impossible or hinders the previous matter: … Sed can introduce something which is surprising if one considers the previous matter: … If the previous words express a denial of something, sed can introduce something which is valid instead: … Sometimes sed shows a contrast with something assumed’)

Wennergren then (p. 305) discusses what he calls “nuanca sed” (‘nuanced sed’):

Nuanca sed ne ligas du aferojn, sed enkondukas frazon, kiu esprimas ŝanĝon de paroltemo aŭ interrompon:

  • Sed ni ne parolos plu pri tiu ĉi punkto. Oni povus imagi subkomprenitan antaŭfrazon: Eblus daŭrigi, sed …

[…]

Nuanca sed povas ankaŭ enkonduki elkrion de surprizo aŭ malkonsento. Tiam oni iafoje povus diri, ke sed kontrastas al la eldiro de alia persono:

  • Mi donos al vi kvin eŭrojn.—Sed tio ne estas justa!

  • Li venkis en la konkurso.—Sed tio estas ja bonega!

(‘Nuanced sed does not connect two matters, but rather introduces a sentence which expresses a change of subject or an introduction:

  • But we shall not speak more concerning this point. One could imagine an understood preceding sentence: One could continue, but …

[…]

Nuanced sed can also introduce an exclamation of surprise or disagreement. Then one could sometimes say that sed contrasts with the statement of another person:

  • I shall give you five euros.—But that is not right!

  • He won in the contest.—But that is indeed excellent!

Wennergren (2005: 300) seems to be aware of the conventional implicature conveyed by kaj ‘and’ that there is some relation between the two clauses that it connects, stating, “Kaj povas ligi tutajn frazojn, kiuj iamaniere kunapartenas” (‘Kaj can connect whole sentences which in some way belong together’). He goes further and gives (p. 301) examples of sentences in which kaj indicates different ways in which clauses “belong together”.

Kiam frazoj estas ligitaj per kaj, tiu ligo povas reprezenti multajn diversajn signiforilatojn. Kia estas la rilato, oni devas kompreni el la kunteksto. Eblas klarigi la rilaton per aldonaj esprimoj:

  • Mi lavis la vestaĵojn, kaj (poste) mi sekigis ilin. Kaj montras tempan sinsekvon. La vorto poste je bezono povas helpi al kompreno.

  • Ŝi aŭdis teruran bruon, kaj (tial) ŝi telefonis al la polico. Kaj montras sekvon aŭ rezulton. Tial povas helpi al kompreno.

  • Anno estas gaja persono, kaj (kontraste) Elizabeto estas silentema. Kaj montras kontraston.

  • Ni klopodis per ĉiuj fortoj, kaj (tamen) ni malsukcesis. Kaj montras neatenditan sekvon. Tamen povas pliklarigi tion. Ankoraŭ pli klara estus sed anstataŭ kaj.

  • Promesu neniam plu fari tian stultaĵon, kaj (tiam) mi helpos al vi reordigi la aferon. Kaj montras kondiĉan rilaton inter la frazoj. Tiam pova helpi al la kompreno. Eĉ pli klare estus transformi la unuan frazon en subfrazon kun se: Se vi promesas neniam plu fari tian stultaĵon, (tiam) mi helpos al vi reordigi la aferon. […]

  • Ŝi havas brunajn okulojn, kaj (aldone) ŝiaj haroj estas longaj. Kaj montras aldonan informon.

  • Li preferis foriri tre frue de la festo, kaj tio ne surprizas min, ĉar vere estis tre enue tie. Kaj montras komenton.

(‘When sentences are joined with kaj, that connection can represent many different meaning relations. What kind of relationship there is can be understood from the context. It is possible to make the relationship clear with additional expressions:

  • I washed the clothes and (afterwards) I dried them. Kaj shows a temporal sequence. The word poste if necessary can help with understanding.

  • She heard a terrible noise and (for that reason) she phoned the police. Kaj shows a consequence or result. Tial can help with understanding.

  • Anna is a cheerful person, and (in constrast) Elizabeth tends to be quiet. Kaj shows a contrast.

  • We endeavored with all our strength, and (nevertheless) we failed. Kaj shows an unintended consequence. Tamen can make it clearer. Still clearer would be sed [‘but’] instead of kaj.

  • Promise never to do that stupid thing again, and (then) I will help you put the affair in order again. Kaj shows a conditional relation between the sentences. Tiam can help with understanding. It would be even clearer if one changed the first sentence into a subordinate clause with se [‘if’]: If you promise never to do that stupid thing again, (then) I will help you put the affair in order again. […]

  • She has brown eyes, and (in addition) her hair is long. Kaj shows additional information.

  • He preferred to leave the party very early, and that does not surprise me, for it was really dull there. Kaj shows a comment.’)

Such detailed accounts of the meanings of a word for ‘and’ (or any conjunction) are very rare in grammars of either natural or artificial languages.

5.2 Interjections

The (type of) meaning contained in interjections has been difficult to describe. Some analysts have claimed that they involve conventional implicatures. Some AL designers describe the meanings and/or functions of interjections, e.g. Ruggles (1829: 58) states, “Interjections serve to express some passion or emotion. […] They are either positive or contrastive”.

Vidal, the designer of the Langue universelle et analytique has an interesting view on interjections (1844: 39):

Le vocabulaire de la langue universelle doit commencer par les interjections, parce qu’elles représentent chacune une pensée tout entière, quelle que soit la nature des signes dont on se sert pour cet effet. L’artifice de la parole consiste à développer ces pensées en séparant l’attribut du sujet par un assemblage de mots que noun nommons proposition.

(‘The vocabulary of the universal language must begin with the interjections, because each of them represents a whole thought, whatever be the nature of the signs which one makes use of for this effect. The contrivance of speech consists in developing these thoughts by separating the attribute from the subject by means of a combination of words that we call a clause.’)

His language has a relatively large number of (words which he calls) interjections; he says (ibid.:40):

Il est des circonstances où il est tres-essentiel de pouvoir exprimer une idée au même instant qu’on la conçoit: si l’on veut, par exemple, appeler au secours dans un danger pressant. J’ai cru devoir agrandir un peu le cadre de celles que nous avons, en observant toutefois que l’on ferait peut-être bien de l’agrandir davantage

(‘There are circumstances when it is very essential to be able to express an idea at the same instant when one conceives it: if one wants, for example, to call for help in pressing danger. I thought that it was necessary to expand the range of those [interjections] that we have, while however observing that one would perhaps do well to expand it more’)

Some of the interjections of his language are equivalent to words or sentences of French, e.g. hol ‘merci’ [‘thank you’], hap ‘que voulez-vous?’ [‘what do you want?’], while others express emotions.

This might lead one to reflect on how to define the class of interjections. If we require interjections to have only non-truth conditional meaning, then many of the words which Vidal calls interjections will not be such. Of course there will be other words which have some non-truth conditional meaning, e.g. but, but perhaps interjections have only this kind of meaning. The question then is whether Vidal and some other language designers are in conflict with linguistic thought about the nature of interjections.

Wennergren (2005: 314) classifies the Esperanto word nu ‘well; now’ among the interjections and says that it has a variety of meanings:

Nu estas ĝenerala atentiga vorto. Ĝi iel montras al la alparolato, ke io speciala sekvos. La precizaj nuancoj estas tre diversaj. Alvoko: “ Nu , mia filino?”—“Jes, patrino.” “ Nu , Alfred,” ekkriis la doktoro, “kion vi diros al tio ĉi?” Malpacienco: Nu , iru pli rapide! Nu, nu malsaĝulo, ĉesu! Konsento, koncedo: Nu , bone, bone! Nu do, venu, se vi vere insistas. Fino de interparolo: Nu , bone do, tial ni iru! Nu , bonan nokton! Surpriziĝo: Nu ! Kiu supozus ion tian? Nu , mi neniam atendis tion! Dubo, hezito: Nu , kiel nun, Anton Antonoviĉ? Klarigo: Nu ! Tion kaŭzis difekto en la hejtilo. Rezignacio: Nu , tia estas la vivo. Nu , kion fari? Konsolo: Nu, nu ! Ne ploru!

(‘Nu is a general attention drawing word. It shows the addressee in some way that something special will follow. Its precise nuances are very diverse. Calling: “ Well , my daughter?”—Yes, mother.” “ Well , Alfred”, exclaimed the doctor, “what will you say to this?” Impatience: Now , go more quickly! Now, now you fool, stop! Agreement, concession: Well , good, good! Well then, come, if you really insist. End of conversation: Well , good then, so let’s go! Well , good night! Surprise: Well ! Who would think such a thing? Well , I never expected that! Doubt, hesitation: Well, what now, Anton Antonovich? Clarification: Well! A defect in the heater caused that. Resignation: Well , that’s life. Well , what can one do? Consolation: Now, now ! Don’t cry!)

Thus, as with his discussion of kaj ‘and’, he gives a more detailed account of the meanings of this word than one will find for the equivalent words in most grammars of natural or artificial languages.

5.3 Illocutionary Force

Utterances with an indirect illocutionary force occur commonly in natural languages. One might therefore expect them to occur in ALs as well. On the other hand, one could see them as a source of possible confusion, like non-literal language (see the next section) and try to forbid their use. To my knowledge, no AL designer has taken such a step. However, what amounts to the same thing has been done in Seaspeak, a controlled language based on English for nautical purposes: Seaspeak has mandatory markers of speech acts. Weeks et al. (1988: 96) (as quoted in Kimbrough and Yang (2005: 303) state:

Maritime messages transmitted over VHF should be short, accurate, and relevant. Furthermore, messages should be transmitted in language simple enough for a non-native speaker of English to comprehend without difficulty.

One useful means of making the language simpler is to indicate, at the beginning of a message, what sort of message it is going to be. Thus, if a question is going to be asked, the speaker simply says the word ‘QUESTION’ before the question itself. Similarly, if a piece of advice is going to be given, the speaker says the word ‘ADVICE’ in advance of his message. There are just seven of these Message Markers and after a little practice, learners should experience no difficulty in using them.

Presumably sentences marked as questions could not be used e.g. as requests in Seaspeak, nor could there be rhetorical questions. However, given the limited domain in which Seaspeak was intended to be used, such indirect speech acts might not occur anyway.

Consider now the following text in Konya (Sulky 2005):

Dialogue 1:

Illustrating the perils of literally translating idiomatic English expressions into Konya:

mon-misi Xenya kesati:

xenye tenwi pofu toku moti yu tufu we xuxin-xeni

Ms. Shen (a visitor):

Can you tell me how to find the train station?

Lukya pan-kesati:

sunye

Luka (a local):

Yep

mon-misi Xenya:

lawa… toku xexenye

Ms. Shen:

Well… tell me, please

Lukya:

sinu mepi yo moti taunwe

Luka:

Look at a map. That’s how

Sulky then comments, “One hopes that in future, Ms. Shen will simply say: xuxin-xeni wa lito kin-lunwi xexenye Where is the train station, please?” He seems to be thinking incorrectly; the question “Can you tell me how to find the train station” is perhaps not what one would call “idiomatic English” but rather involves an indirect speech act, a request in the form of a question. If we take Sulky’s remark in a general sense, then he is saying that there should not be indirect speech acts in Konya. However, this could make Konya unsual when compared with natural languages since one might think that indirect speech acts occur in all of the latter; according to Brown and Levinson (1987: 142), “indirect speech acts are universal and for the most part are probably constructed in essentially similar ways in all languages”.

Indirect speech acts can be found at least some ALs; below are some (possible) examples:

Ido:

Kad vu voluntus pasigar la pano?

‘May I trouble you for the bread?’ (ApGawain et al. 2008: 34)

Here (from a group of example sentences, i.e. not one that actually occurred in speech) we have a question functioning as a request. For some reason, ApGawain’s English version is not literal; a more literal version is ‘Would you be willing to pass the bread?’ (the English version is given before the Ido version).

Eurolengo [from a simulated “typical business letter” (Jones 1972: 63)]:

nos gustaral resevar sampels and pryses for noster consideration

‘we would be pleased to receive samples and prices for our consideration’ (ibid.:62, tr. 63)

This statement could be seen as an indirect request.

Interlingua (IALA) Footnote 8 The context is from a pseudo-dialogue involving a man, his son, and another man. The boy asks his father what the other man is doing; apparently he is writing a letter:

Johnny (al senior): Senior, scribe vos un littera a vostre matre?

‘Johnny (to the gentleman): Sir, are you writing a letter to your mother?’

Le senior (in un tono un pacuo irritate):

No!

‘The gentleman (in a tone a little irritated):

No!’

Johnny:

Que face vos alora?

‘Johnny:

What are you doing then?’

Le senior:

Io attende

‘The gentleman:

I am waiting.’

Johnny:

Que attende vos?

‘Johnny:

What are you waiting for?’

Le senior:

Io attende un momento de silentio pro finir iste littera

‘The gentleman:

I am waiting for a moment of silence to finish this letter.’

Johnny:

Papa, ille attende un momento de silentio

‘Johnny:

Papa, he is waiting for a moment of silence.’

Senior Smith (con un profundo suspiro):

Ah, si. Multe gente attende un momento de silentio

‘Mr. Smith (with a deep sigh):

Ah, yes. Many people wait for a moment of silence.’

Johnny:

E nos, que attende nos? Nos non attende un momento de silentio, nonne?

‘Johnny:

And us, papa, what are we waiting for? We aren’t waiting for a moment of silence, are we?’

Senior Smith:

No, Johnny, nos attende mama!

‘Mr. Smith:

No, Johnny, we are waiting for mama!’ (Gode 1954: 41)

One could interpret the gentleman’s statement that he is waiting for silence as an indirect request for silence. Johnny fails to understand this, perhaps due to the fact that he is a child (one should bear in mind that this passage is meant to be humorous, and without Johnny’s misunderstanding it would not be so), and his father does not enlighten him. (The book in which this appears, Gode (1954), is based at least in part on a book for learning Spanish, and so this psedo-dialogue may originally have appeared in Spanish or English; one might think that the indirect speech act occurs here for that reason. However, if such speech acts were not permitted or were not possible in Interlingua, presumably this pseudo-dialogue would not have occurred in Gode (1954).)

6 Metaphor and Non-literal Language

Although metaphors, and non-literal language more generally, are a significant part of every natural language, some ALs designers frown on their use and try to eliminate them from their languages. Consider the remarks of Morneau (2006: sec. 27.0) (Morneau is the creator of Latejami):

when speakers of natural languages use non-literal language it is almost always because they are forced to do so. They cannot avoid it either because their vocabulary does not have an appropriate literal construction available, or because it is something that the speaker is not comfortable using.

This is unfortunate because the way that a non-literal construction will be interpreted will depend very much on the native language and culture of the listener. For example, metaphoric use of the word “pig” can have meanings such as “slob”, “sex maniac”, or “over-eater” in English, but will have different meanings to speakers of other languages. Also, as we’ve seen many times throughout this monograph, many metaphors, including the above examples, can be avoided by using appropriate derivations instead. For example, pejorative morphemes or more precisely derived compounds can be used to implement the above examples. In fact, I have become completely convinced that a properly derived word can replace any required or unavoidable metaphor, and it can never be misinterpreted by native speakers of other languages.

The goal of a designer of an MT interlingua should be to provide the means to say anything without the need for non-literal language. In other words, metaphor, polysemy, and idiom should be optional—they should never be obligatory. It is also my opinion that non-literal language should be generally avoided (except where its use is obvious to all listeners or readers), since the possibility for misunderstanding is so great.

Morneau’s statement that non-literal language is not employed by choice seems to be incorrect. Other AL designers share his negative view of non-literal language: in his work on Hom-idyomo Cárdenas (1923: 153) says:

Words should be used and interpreted in their natural meanings and not with the figurative meanings they may have in other languages. Thus, pesto [‘pest’] should not be used in the sense of “invective”, nor maro [‘sea’] in that of “abundance”, nor nigra [‘black’] in that of “sad” or “gloomy”. The expression, Gladyo sitya di sango [‘sword thisty for blood’] may be very poetical, but it is not true. A language which has not come into general use should not be employed figurative, at least for the translation of idioms and saying. That will come later.

What is probably behind such views is a desire for clarity and ease of understanding in a language. A way of preserving this while permitting non-literal language is to have an indicator of non-literal language, and several ALs have such a device. One of these is Sotos Ochando’s Lengua Universal (an a priori language). Gisbert and Lorrio in their (1862) pedagogical grammar of this AL express a more tolerant view of non-literal language: (pp. 76–77):

Se ha sentado por base que las palabras de la Lengua Universal no han de tener doble sentido, en lo cual consiste una de sus principales ventajas; pero debe comprenderse que al decir que ninguna palabra tendrá doble sentido, no excluimos ni podemos excluir las figuras que como naturales al hombre, y procedentes de su mismo espíritu, no hay nadie que pueda proscribirlas de la locucion. Su usa nunca induce á error, como acontece con las palabras que sin usarse figuradamente tienen sentidos diversos, representando objetos distintos. Cuando de un Papa se dice que lleva el timon de la nave de la Iglesia, nadie se confunde, nadie toma á la Iglesia por un buque: quando se dice, hay nubes en el horizonte politico, nadie mira al cielo á ver si está raso. […]—En este Lengua subsistirán por consiguiente las metáforas y las figures todas del pensamiento, y subsistirian aunque la voluntad del autor fuera excluirlas; pues son efecto natural de la imaginacion que en ellas busca la explicacion mas viva y pintoresca de sus ideas, y que sin sablerlo las usa (de continuo) aun en el lenguaje vulgar, principalmente al hallarse afectada de una pasion cualquiera.

(‘It has been stipulated that the words of the Universal Language should not have double meanings; in this consists one of its main advantages; however, this should not be understood to mean that no word will have a double meaning: we do not exclude, nor can we exclude figures [of speech]; as they are natural to man, and proceed from his very spirit, there is no one who could proscribe them from speech. Their use never leads to error, as it does not with words which, without being used figuratively, have different meanings, representing different objects. When someone says of a Pope that “he takes the helm of the ship of the Church”, no one takes the Church to be a sailing vessel; when someone says, “there are clouds on the political horizon”, no one looks at the sky to see whether it is clear. In this Language consequently there will still be metaphors and all the figures of throught, and they would exist even if it were the will of the author to exclude them, since they are a natural effect of the imagination, which seeks in them the most vivid and picturesque expression of its ideas, and which, without knowing it, uses them (continuously) even in everyday language, mainly when affected by some passion.’)

Oddly enough, in a work by Sotos Ochando himself (1863) there is a more restricted view of non-literal language (p. 33):

En la Lengua Universal no se admiten figuras de letras ni de diccion. Solo se exceptúan las licencias esplicadas sobre esto, porque están tan fijas sus reglas, que no dejan lugar á dudas ni equivocaiones. Sin embargo, es conveniente y aun necesario admitir tres clases de figuras ó metáforas de sentido.

  1. 1.

    a La sustitucion de ciertas voces, cuando la reclaman la moral, la decencia, el decoro ó la delicadeza. Tales son las que significan las partes pudendas de ambos sexos, varios de sus actos, los objetos que excitan asco, etc., etc.

  2. 2.

    a La metáforas en que se toma un objeto, una cualidad, una accion por otra, atendida la relacion que tienen entre sí los objetos por su semejanza, participacion ú otra causa, v.g., cuando decimos la aurora de las ciencias, el azote de Dios.

  3. 3.

    a Las figures de retórica, á lo menos muchas de ellas, como la hyperbole, la ironia, la personificacion.

En efecto, estas tres clases de figures están fundadas en la naturaleza del hombre, y con mas ó menos extension son communes á todos los tiempos y paises.

(‘In the Universal Language figurative language is allowed neither in written nor in spoken language. The only exceptions are the liberties explained concerning it, because their rules are fixed to such an extent that they do not allow room for doubts or errors. However, it is desirable and even necessary to allow three kinds of figures or metaphors of meaning.

  • 1st: Substitution for certain words, when morality, decency, decorum, or delicacy calls for it. Such [words] are those which signify the private parts of both sexes, various of their acts, things which arouse disgust, etc., etc.

  • 2nd: Metaphors in which a thing, a quality, an action is taken for another one, on the basis of a relationship which holds between the things because of their similarity, participation [in the same action?], or another reason, e.g. when we say the dawn of the sciences, the scourge of God.

  • 3rd: Rhetorical figures, at least many of them, such as hyperbole, irony, personification.

Indeed, these three kinds of figures are grounded in the nature of man and to a greater or less extent are common to all times and countries.’)

The indicators mentioned above are not always required in Sotos Ochando’s language, nor are those of another type, those which mark the technical use of a word; among the “licencias” (‘liberties’) mentioned by Gisbert and Lorrio (1862: 94–5) is the following (p. 95):

Pueden tambien suprimirse muchas veces los signos del sentido metafórico; pues que este se conoce con frecuencia por el contexto, como sucede en las demas lenguas, que no tienen otro medio de indicarlo. Esta supresion será muy conveniente en algunos casos, para que aparezca la fuerza y belleza de la metáfora.

Las mismas observaciones pueden aplicarse al uso de los signos del sentido técnico.

(‘Many times the markers of metaphorical meaning can also be omitted, since this [meaning] is often known by the context, as happens in other languages, which do not have any other means of indicating it. This omitting will be very desirable in some cases, in order that the strength and beauty of the metaphor appear.

The same observations can apply to the use of the markers of technical meaning.’)

In addition to these words Sotos Ochando’s Lengua Universal has a set of indicators which seem to be heterogeneous; Sotos Ochando (1863: 37) labels them “[v]oces significativas de un sentido especial de la frase” (‘words indicating a special meaning of the sentence’); the list of those that he “proposes” is (ibid.):

Ar :

para sentido interrogativo (‘for interrogative meaning’)

Er :

para el dubitativo (‘for dubitative meaning’)

Ir :

para el irónico (‘for ironic meaning’)

Or :

para el admirativo (‘for admirative meaning’)

Ur :

para el de sorpresa (‘for surprise’)

As :

para el de sentido optativo (‘for optative meaning’)

Es :

para el de indignacion (‘for indignation’)

Is :

para el depreciativo (‘for depreciative’)

I say that they may be heterogeneous because, at least in natural languages, their equivalents would not all be of the same type. Some of them might be equvalent to mood/modality markers, e.g. the Greek optivative mood endings. However, this would not be true of e.g. ir or es; I do not know of any language that has an “ironic mood” or an “indignant mood”. With the possible exception of ar (the only one of these markers that I have found in any texts in the language), I do not believe that any of these words are markers of illocutionary force along the lines of those in Seaspeak. (I am not certain whether even ar is such a marker, since it is not clear whether a sentence marked with it could have an indirect illocutionary force of something other than a question.Footnote 9) I would hesitate to classify words such as ir and es as interjections, although they express emotions, since, if they act syntactically like ar, they are part of a sentence, while one feature attributed to interjections is their lack of connection with other syntactic units. (It is difficult to determine whether all the words of this group act in the same way, since, as I have noted, ar is the only one that I have seen in context.)

Sotos Ochando (ibid.) justifies the creation of these words as follows: “Aunque el tono debe acompañar y caracterizar estos sentidos, es convenientísimo que haya palabras bien fijas y determinadas que excluyan toda duda y equivocacion.” (‘Even if the tone [i.e. intonation] must accompany and characterize these meanings, it is most advantageous that there be very fixed and specified words that would eliminate all doubt and error.’) He also says, “Esta clase de modificativos son frecuentes en todas las lenguas” (‘This type of modifiers is frequent in all languages’).

Glosa is another AL which is not friendly towards figurative language, and, like Sotos Ochando’s Lengua Universal, there is a marker for it, or rather, such a marker is recommended; Sect. 8.vii of Gaskell (1999) is called “Idiom and metaphor” and it reads as follows:

For clarity of expression across cultural boundaries these should be avoided in Auxiliary Language usage; however, where it is necessary—for literary purposes—to quote a national-language idiom within Glosa, then such non-literal language ought to be marked with some ‘neutral’ symbol, EG pluvi ^plu feli e kani^ [‘rain ^cats and dogs^’].

Grzega (2005: 67–8) says about Basic Global English, a simplified version of English which could be used both for pedigogical purposes and as an international auxiliary language, that “Native and advanced non-native speakers of English are asked to… abstain from metaphorical expressions that cannot be interpreted word-for-word (as these have shown to be problematic in lingua-franca communication)—in this respect a certain awareness competence might have to be trained”. In a later paper (Grzega 2008) he states (p. 140), “Metaphors should only be used if objectively obvious and should be marked (this is like…)”. That is, Basic Global English, like Sotos Ochando’s Lengua Universal and Glosa, has an overt means of marking metaphors.

Those language designers who try to exclude the possibility of using non-literal language in their ALs may be attempting something which would make their languages unnatural, at least if they are languages intended for the same general purposes as natural languages, although this does not apply to all artificial languages, if we take “artificial languages” in a broad sense; consider the following remarks by Cohen (1993: 59):

it is clearly characteristic of certain categories of artificial languages that they must lack any possibility of metaphor. Programming-languages for computers, like Fortran, or interpreted formal systems, like Carnap’s, would be very seriously flawed in the performance of the tasks for which they are severally designed if they allowed their component words or symbols to be attributed new and unstipulated meanings in certain contexts. It is arguable, therefore, that we radically blur the difference between these kinds of artificial languages on the one hand, and ordinary natural languages, like English (or artificial languages for everyday use, like Esperanto), on the other, if we do not allow essentially for the possibility of metaphor in our analysis of the latter.

Metaphors and non-literal language do occur (without being marked as such) in at least some ALs. Consider what Gledhill (2000: 122) says about Esperanto:

Many expressions are used non-literally in Esperanto. For example, forpasi ‘to pass away’ = to die, zumi ‘to buzz’ = to potter about, celi ‘to aim’ = to get at/mean something. Some expressions, especially compounds, involve a non-literal sense derived from the donor languages (for example, librotenado from English ‘bookkeeping’) or have emerged because of generally expressed euphemisms (necesejo ‘the necessary plac’’: the toilet). […]

Proverbs and clichés are complex and very fixed forms of metaphor. They generally involve truth-statements and have an element of word play about them. In the Fundamento Zamenhof equipped the language with a large number of proverbs… […]

Although most proverbs are rarely invoked in the general language, they form part of the basic repertoire of metaphors which the speaker may allude to (fera mano ‘an iron hand’, amata ĉevaleto ‘hobby horse’). By writing down a set of proverbs, Zamenhof effectively created an oral history of the language, a corpus of expres[s]ions to dip into and cite or reformulate.

Note also the existence of the book Dahlenburg (2006), whose subtitle is Stilfiguoroj en la poezio de esperanto (‘Figures of style in Esperanto poetry’); among the figures included are metaphor, irony, and euphemism.

7 Texts and Comments

I now present some texts and examine them from a pragmatic point of view. I have already mentioned the fact that ALs are generally not native languages, and thus the pragmatics in them may be affected by a speaker’s first language. Such a carry-over may also occur in AL texts which are translations of texts in another language, so for pragmatic analysis it is probably better to look at original texts in an AL (unless one wants to see whether there are any pragmatic differences between the original and the AL translation). Unfortunately from this point of view, many texts in ALs are translations of well-known works in natural languages. Even if a text is not obviously a translation, or stated to be one, it may well be one: given that even creators of ALs have another language as their first language, many or most texts which they write in their ALs may be translations from their first language. That is, they may usually think of or formulate a sentence in their native language and then give its equivalent in their AL.

It may therefore be almost impossible to find texts in ALs which are not translations in some sense, with the exception of those in the major ALs such as Esperanto which some speakers have been using extensively for decades (and are thus able to think in, i.e. to use without first translating in their minds from another language), and these major ALs may be of less interest here because they are a posteriori. However, we can at least avoid the texts that we know to be translations (unless we are specifically interested in the possible transfer of pragmatic features in translation).

It may also seem best to analyze conversations or dialogues rather than passages involving only a single participant, and many works on ALs contain what appear to be conversations (given in written form). However, they are not real conversations/dialogues in the AL, since they were created by a single person (the author of the work); rather, they are presentations of what a conversation in the language (on a given topic) would or should be like. As noted above, one could assert that they still represent a type of language use, just not the use that they appear to be (i.e. conversation). There may be real dialogues in some a priori and mixed ALs, for example postings and replies to them in internet sites devoted to these languages, but these may often be by people who are not completely fluent (or even close to fluent) in the AL, and thus their pragmatic (and other) features may be taken from their first language. A further problem with some AL texts is that it is not always clear whether a sentence is meant to be part of the same conversation as the previous sentence, or indeed whether a conversation is being portrayed at all.

I have chosen for analysis a text from the a priori language aUI which contains an apparent dialogue. However, it is not even a pseudo-dialogue of the sort just described, but a dialogue occurring within a fictional piece, one of the “[r]eadings” contained in Weilgart (1979). Nevertheless, it may be of particular interest because the participants in it are a human child and an alien who is partly animal and partly plant; if there were a situation where one might expect unusual pragmatics, this might be it. First I give the context only in English translation:

Ever since the space-man had visited Johnny, the boy had envied the little animal-plant of quiet mind; for this creature could travel through infinite space. Why could not Johnny return the visit and come to the distant planet of the star in outer space? The spaceman had told Johnny that in machines and inventions the earthmen were just as good as the space-men (Weilgart 1979: 165)

Below is the dialogue that follows this passage:

1. –“fUd fnu tykwerv a!”, DJoN nIpaV.

2. “Then let us conquer space!”, cried Johnny.

3. – yUg, a-u tygrOpAv: ‘bum UI sEfU: “tykwe” Ub a, vEv fE tykwe y-twam rUt bnu. rUt-A-jAg bnu cEv jOm bnu cEv, bnu yA watAv tykwev a. pfE ć, bnu yA watAv dav ad nEn Oki akiA Ud bUt ne Ub i yt a “bijE-da” tygle at yf. fnu a-u, xnu dav fUd, yc fUIv sE: “tykwev” a. –

(3) But the spaceman had smiled: “Your very word ‘conquest’ of space makes this conquest impossible for you. As long as you are as you are, you will never be able to conquer space. That is, you will never be able to travel through thousands of years with almost the speed of light from one ‘milky-way’ spiral to the next. Now we space-men who do this do not call it ‘conquering’ space.

(4)– Fnu rykOmQ UIv: agtev a, gaf, tykev tag a; yUt ag-niO-Uj bu tyv at retgUv ypums Ub a, bu pI nEk iUv, hU u yc wav tykwev sE. am KU tykwev a. yUg fnu cEv yn-ynam ayn Ub knynE Ib, rUt fnu, a cEv y-tnak-wam.

4. We say modestly: entering space, or, submerging in space; for the more you get to learn the mystery of space, the more you understand why man cannot conquer it. Only God conquers space. But we are tiny specks of dust and, for us, space is unlimited.”

II) dvU, hUd au av. (The Spaceman’s method).

1. – ynDJoN tykOm tygrOpApAv at au: fu yc bav Otgu rUt bum nUm atiO Ib nykam U-gUw. yUg fu tOv at gUv, hUd bnu Ev sE: hUd bnu wav tAv avAm rUt Oki akiA yb tyv iEv-do rUt ves Ib od rUt bnufU?”

1. John had scoffed at the Spaceman: “I am not interested in your world-view and deep philosophy, space-man. But I want to know how you do it. How can you go on for thousands of years without refueling and without eating?”

– au vUtsepAv: bu OtgUv rUt fnum da Ib Ed, yUg bnum da Ub o yc fnum o. jUf, Qg fu wav-yEc UIv fE at bu, fu yc tOv-yEc sE.’

The spaceman decided: “you are interested in the ways and means but your way of life is not our life. Even if I could tell you I would not want to.”

– ynDJoN hIpAv vufU: fnum uamA vEv hUm yjU, hI? “bu UIrv at fu: hUd u wav kad-ov ad Oki akiA Ib krOv tykwe ek a Ib A?”

Johnny wondered: What difference does our attitude make? “Tell me how can you survive thousands of years, and triumph over space and time?”

Au: ‘jUf, Qg fu UIv-yEc fE, bu yc iUv-vEc fu. Bu ova g yga-da Ub o Ub yrkO. Nykam tyk ag gaz, fnu Utev ryko: fnum ypus cEv rykO.’

The Spaceman: “Even if I told you, you would not understand me: You live the surface way-of-life of pride. Deep down in the center we find humility: Our secret is humility.”

2. —ynDJoN krOIpAv: “bu c’krOIyv-wam eb bum rOkU-Uis.”

2. Johnny laughed: “You are ridiculous with your morality.”

—au cpA yktrUm: ‘yUt fu yc wav typev at fnum ki, yUt fu sepAv fum o, Ut ytyrAv bnum eki, fu tEvAm rUt fum banu, tyg-ytwepAm-s, sE yc pwUrm, hE bu gaf ym-u Uv UI fu. Fu UItAv fE at bu: bu ymA AiOpAv tok, hI?’

The Spaceman was serious: “Since I cannot return to our stars, since I gave my life to save your planet, becoming an outcast to my tribe, it does not matter what you or anybody thinks of me. I will tell you: Have you ever watched a tree?”

3. —ynDJoN tEpAv y-trAwm: “fu gUv nEn tok.”

3. Johnny grew impatient: “I know many trees.”

—‘yt hE tok tnev?’

“Out of what does a tree grow?”

—“yt to.”

“Out of a seed.”

—‘to cEv Uj nam Uj tok?’

“Is the seed as big as the tree?”

—“yr! to c yn-ynam.”

“No. The seed is tiny.”

—‘fA, xA to tnev tag tok, to yc tyv namU, hI?’

“Now, when the seed grows into the tree, does it not gain greatness?” (ibid.:165–8)

The first comment one might make is that there is nothing very strange here from a pragmatic point of view, in spite of the unusual context and the odd appearance of the language (and aUI does seem to be unlike natural languages and many ALs in one way: it is meant to be (largely) iconic, with there being a relation between the sound of many (components of) words and what they mean).

Johnny appears to be rude (and uses rather direct language), but that could be attributed to him being a child and/or being an arrogant earthman. However, to really know the level of impoliteness, one would have to have a sense of the exact meaning, including connotations, of krOIyv-wam ‘ridiculous’, which is impossible without analysis of many other occurrences of it (if there even are such). It is composed of krOIv ‘to laugh’ and wam ‘able’, and thus means ‘able to be laughed at’, but it is not clear whether it has the same sort of negative connotations as English ‘laughable’; one does not always laugh at something/someone because of its/perceived negative qualities.

Next we will see a pseudo-conversation in another a priori language, Sotos Ochando’s Lengua Universal (Gisbert and Lorrio 1862: 121; I have translated their Spanish translations into English):

  • Ar saban labli riolarbem?

  • ‘What family do you have?’

  • Riolarbel le sacan lasfie, siba lalcae bal sibi leldes.

  • ‘I have my wife, one daughter and three servants.’

  • Be saban afaca riaburben sacen lague?

  • ‘Where is your mother?’

  • Sacan lague riaburben soreboc fle sacan lamee: mu sodibi agoldirbin glo sacan imari bal riaburbin fle sacas.

  • ‘My mother is now with my sister: in summer she will come home and will be with us.’

  • Riabirben ibelin sacen lague?

  • ‘Is your mother very old?’

  • Nan riabirben ibelin: riolarben sicra sugas bal riaburben gan ipafon.

  • ‘She is not very old: she is 60 years old and very well.’

  • Bal sacen lalcae ar le saban sugas riolarben?

  • ‘And your daughter, how old is she?’

  • Riolarben le sibra sugas.

  • ‘She is six years old’

  • Bal sace nan obamerbem se lasur?

  • ‘And you, aren’t you thinking of getting married?’Footnote 10

  • Riabirbel gan ibefon mal se lasurdel.

  • ‘I am very young to get married.’

  • Riolarbem sicebu sugas: se lasurfom sorogoc: sace riabirbim ol ugefon lasfi.

  • ‘You are 25 years old: get married soon: you will be a good husband.’

Again there is nothing very strange here, although some parts of the conversation involve more directness than one would see in some natural languages; in English it might not be considered to ask whether someone’s mother was “very old” or to tell someone to get married, unless one knew him well (which would appear not to be the case in this pseudo-conversation).

Let us now look at part of a pseudo-conversation in Hom-idyomo, an a posteriori language (Cárdenas 1923: 66–7; note that it may not always be clear which interlocutor is speaking):

Irivi nos a dinerir, gesseñros, bite.

‘Let us go to dine, gentleman, if you please.’

[…]

Señoro Franklin, degnivi sedentir be en cia sejo.

‘Mr. Franklin, please take this place.’

Danke.

‘Thank you.’

Kay bi, señórino Danton, bite sendenti be tie.

‘And you, Mrs. Danton, please sit here.’

Permitivi me, demandir be la menuo.

‘Allow me to ask you for the bill of fare.’

Veri volunte, madamo.

‘With pleasure, madam.’

Favorvimi per la pano.

‘Oblige me with the bread.’

Mi rogay be pasir la súpero.

‘Pray, pass me the tureen.’

Voluntay bi miksir raspata-kaseo kun la supo?

‘Will you have grated cheese with your soup?’

Danke, after bi.

‘Thanks, after you.’

Preferay perhapse las gesseñros ke oni apririn las fenestros? La ambyento komencay a devenir veri kalora.

‘Perhaps you gentlemen would prefer to have the windows open? The weather is beginning to be very warm.’

Kiel bi voluntin, madamo. Nos no sensacyonay multa kaloro, sed bia gardeno estay tante bela ke mirir je estay, en everya cirkunstanco, una plesuro.

‘As you like, madam. We are not very warm, but your garden is so beautiful that it is always a pleasure to look at it.’

[…]

Gustay bi la karno sanga?

‘Do you like your meat underdone?’

Mi preferay je koktata.

‘I prefer it well done.’

Ambi klasos havay estite koktatas, la una plus dan la otra.

‘Both kinds have been cooked, the one more than the other.’

Perhapse bi volutay dirir ke ambi estey kaloratas per fayro, una til kwande ji koktey kay la otra nur til kwande ji komencey a emanir una likido sángis̃a, remanante kruda interne.

‘Perhaps you mean that they were both heated over the fire, the one until it was cooked, and the other until a bloodlike liquid oozes out, the inside remaining raw.’

Permitay bi me, señoro Pasteur, demandir bia opino pri cia temo?

‘May I ask your opinion about the subject, Mr. Pasteur?’

Veri plesure, madamo. Si mi no eroray, la karno sanga kontenay ankora mikrobyos danjeras por la saluro, kay ji no estay plus alimenta dan cia wel koktata. No pensay bi mesme, señoro Franklin?

‘With great pleasure, madam. If I am not mistaken, underdone meat still contains microbes dangerous to health, and is not more nourishing than that which is thoroughly cooked. Do you not think so, Mr. Franklin?’

Mi no estay kompetenta kiel bi pri cia topiko, sed mi imaginay ke mandukir karno insufice koktata estay una kapriyo de la modo. Kio mi posiblay an informir be estay ke mi no gustay je.

‘I am not competent in this matter, as you are, but I imagine that eating insufficiently cooked meat is a whim of fashion. All I can tell you is that I do not like it.’

This also does not appear to be a particularly unusual conversation although it goes from what is, by contemporary standards, a high level of politeness to a distasteful remark, which one might not expect to hear at the dinner table in Cárdenas’ time or ours.

The last text that I shall present is in Esata (Bothi 2006:Dlog fav (‘Dialogue five’))Footnote 11:

A::

Va kand cu yufila cuha?

‘What kind of food do you feel like eating?’

B::

Yola cinesi, nayu?

‘I like Chinese, and you?’

A::

Mitu, benotu hofi, na jelasi satd hada so

‘Me too, but not too often, and just last Saturday I had some.’

B::

Derz de mesiki ples raqi, wona trayahe?

‘There’s the Mexican place right here, want to try it?’

A::

Wuno mayna be yo vewe yita tumuc mesiki, den noslipa gu

‘Wouldn’t mind but I always eat too much Mexican, then don’t sleep well.’

B::

Wel wikugo tode sifud resteran. Fila yita sifud?

‘Well we could go to the seafood restaurant. Feel like eating seafood?’

A::

Yokugo fone bekda fix plet, dazgu nano kosa tumuc. Hobotyu?

‘I could go for a baked fish plate, that’s good and doesn’t cost too much. How about you?’

B::

Yola de mixi sifud platr wit sofa hevte, tune gudil

‘I like the mixed seafood platter with some of everything, also a good deal.’

A::

Hok, hez sifud den, legohina. lediz fersi

‘Ok, it’s seafood then, let’s go in. Ladies first.’

B::

Hune lediy? Yobinewu, nane rilwan tu!

‘Who’s a lady? I’m a woman, and a real one too!’

A::

Yubigimi negran haptit. Wona sita qi?

‘You’re giving me a grand appetite. Want to sit here?’

B::

Das fayni, jenosita woyukesi detivu

‘That’s fine, just don’t sit where you can see the TV.’

Once again there is nothing exotic here. There is some indirectness. For example, in answer to B’s question of whether he likes Chinese food, which really is asking whether he wants to have Chinese food on this occasion, A does not say “no”; rather he says that he does not like it (although “not too often”) but had it recently. When asked about having Mexican food, A again does not say that he does not want to, in fact he says that he “wouldn’t mind” but indicates that he does not favor this possibility. (In fact, it appears that he does like Mexican food, otherwise he would not “always eat too much” of it.)

8 Conclusion

We have looked at ALs with respect to several areas of pragmatics. Although some ALs, particularly the a priori ones, on the surface seem to be quite strange, in general they are not very odd in terms of pragmatics. A small number of them have features which are rare or non-existent in natural languages, e.g. the overt markers of illocutionary force of Seaspeak and the markers of metaphors of Sotos Ochando’s Lengua Universal.

The relatively small number of remarks and instructions concerning pragmatics in works on ALs indicates that designers of ALs and authors of books about them have generally not thought much about this area of their languages. This should not be surprising; textbooks on natural languages often do not mention pragmatic matters, with the exception of brief remarks about some aspects of polite language, e.g. formal and informal pronouns. Thus it would be very unusual for a French textbook to discuss conversational implicatures. One could argue that this is because conversational implicatures are a universal features of languages and so do not need to be discussed or explained for those learning e.g. French.

On the other hand, it seems that conventional implicatures may differ to some extent among natural languages, at least with certain items; while in all languages words meaning ‘and’ probably have the conventional implicature that the clauses connected with them have some sort of relation (unlike the & of propositional logic), and it would be hard to imagine that it could be otherwise (what would be the point of connecting two clauses if there were no relation between them?), there may be differences involving words for ‘but’, namely in the strength of the contrast that they signal. For example, the Russian word a signals a weaker contrast than English but. In spite of such differences among natural languages, works on ALs rarely give information about conventional implicatures of conjunctions.

One could argue that from a pragmatic point of view ALs are interesting because they are not interesting; that is, as exotic as they may seem on the surface, and as exotic as they may be in certain respects (in particular, the way in which the lexicon has been constructed in some a priori ALs), their pragmatics are often not significantly different from those of natural languages. This is perhaps because ALs simply have carried over the pragmatics of one or more natural languages, even if they are supposedly a priori; perhaps it is quite difficult to create a truly a prior pragmatics, at least if one is trying to design a usable language.

This brings us to the reason why the lack of strangeness of AL pragmatics might be interesting. If there are pragmatic universals, that is, if the pragmatics of language is wired into the human brain (as has been claimed for various syntactic principles), then this might limit not only competence but also language creation (unless one were trying to be perverse). I have made this point before (Libert 2000: 1) with regard to universals in general, but the argument might be even stronger with regard to pragmatics

If a language were designed with a pragmatic feature that violated some universal, i.e. if it had an “unnatural” feature, one might expect that, if it were used for long enough by enough people and were not limited in the domains that it occurred in, i.e. if it fulfilled all the roles that that natural languages do (e.g. conversation, literature), this feature might disappear. For example, one might think that markers which clearly and overtly indicated illocutionary force, and which therefore prevented the existence of indirect illocutionary force, could not exist in a natural language. Of course many languages have ways of marking e.g. questions, but perhaps in all such languages questions can have an indirect illocutionary force as something other than a question (such as a command). It would not be surprising if eventually users of a language with Seaspeak-type markers started using sentences with question markers with some indirect illocutionary force. This will probably not happen with Seaspeak due to the very limited contexts in which it is supposed to be used, but one could imagine the process taking place with more widely used languages. Thus, an unnatural pragmatic feature might only be able to survive in a language which was designed for particular and narrow functions.

In any case one might ask what the point of creating a new pragmatics for an AL would be. ALs have often been created in an effort to improve and/or simplify natural languages, involving e.g. the elimination of irregular verbs, and such features may have been criticized by AL designers. If these designers had felt that some pragmatic features of a natural language (or natural language in genera) were better removed, then they could have done it, or at least attempted to do this. Indeed this has been done, as we have seen, with respect to metaphor and the familiar-polite pronoun distinction (recall Zamenhof’s attitude toward ci). However, to my knowledge, no AL designer has argued for changes concerning conversational implicatures, presuppositions, or conventional implicatures, with respect to particular items in a language or in general.