metaphorik.de 01/2001 - Radden, The folk model of language

The notion of ‘language’ is not a primary concept but tends to be derived from more basic notions within the language frame. This paper looks at the ways notions of articulation are metonymically used in different languages to stand for ‘speaking’ and ‘language.’ It is assumed that these unidirectional metonymic shifts and their metaphoric elaborations reflect a naive view of language, which may be seen as revealing a ”folk model of language.” In this simplified model, our understanding of speaking and language is reduced to articulation and a small set of speech organs. Thus, the throat and voice are associated with meaningful speech, the tongue is commonly seen as the generator of a word’s sounds and meaning and, as the most salient articulator, tends to be associated with the notion of ‘language,’ the mouth as the second most important articulator may also contribute to a word’s meaning, and the teeth and the lips may, as the final articulators in the production of speech, distort a word’s meaning.

0. Introduction

The notion of ‘language’ is not a primary concept in the sense of referring to a ”fundamental aspect of our cognitive machinery” (see Grady 1997). There do not appear to be many languages that have a word exclusively reserved to denote language.[1] As a rule the word for language is synchronically related to, or historically derived from, a more basic sense. These earlier or basic senses tend to belong to one of the following domains: (i) articulation and speech organs, (ii) linguistic action, and (iii) basic linguistic units. Expressions denoting articulation such as voice or speech organs such as tongue are often used to describe modes of speaking and language. Expressions denoting linguistic action such as speak are also commonly used in reference to language. In many languages the word for language derives from a word meaning ‘speak, say, tell.’ For example, Latin Çr~tiÇ derives from Çr~re ‘speak formally,’ Irish urlabra from labrar ‘speak,’ Old Norse m~l from mFla ‘speak’; similarly Dutch taal, German Sprache, etc.[2] Expressions denoting basic linguistic units such as word are less frequently used in the sense of ‘language.’ Examples include Tagalog salit> ‘word, language,’ Japanese kotaba ‘word, language,’ and Basque hizkuntza, which is composed of hitz ‘word’ and the noun-forming suffix –kuntza and may thus be glossed as something like ‘word-activity.’[3]

The semantic shifts characterizing all these examples are metonymic: they operate within the same conceptual frame, which might be described as ”language frame” or ”idealized cognitive model of language” (ICM). These metonymic shifts are found across languages and are certainly not haphazard but cognitively motivated. They allow us to come to grips with the abstract notion of language by means of more concrete experiences within the language frame. This applies especially to the use of ‘speech organs’ for ‘language,’ which is the subject of this paper. As will be shown below, the metonymies for speaking and language rely on a small set of speech organs only, in particular on those which are tangibly involved in articulation. These metonymies tend to be elaborated by metaphor, for which Goossens (1995) coined the term metaphtonymy. For example, in the expression to be tongue-tied, tongue metonymically refers to ‘ability to speak’ and its being tied is metaphorically understood as ‘inability to speak.’ It is argued that these metonymies and implicatures derived from them as well as metaphtonymies elaborating the metonymies reflect our naive understanding of language, which might be described as a folk model of language. The notion of folk model is understood here in the sense of a blatantly false cultural model.[4]

In order to gain the widest possible picture of the use of metonymies and metaphors for ‘language’ and, ultimately, as evidence for a folk model of language, examples were collected from various, randomly chosen languages. All the examples listed below were provided or verified by native speakers or non-native informants who are fully competent in the particular language.[5] These examples will certainly bring to mind more examples in these or other languages, confirming or modifying the view of the folk model presented in this paper. I should gladly appreciate any response to the folk model developed here.

The main interest at this exploratory stage of the study is in finding cross-linguistic evidence for a folk model of language, which, to my knowledge, has thus far not been the subject of a linguistic or psychological investigation. Possibly the ways we speak about language are very similar cross-linguistically, and the folk model reflected in our metonymic use of language is universal. Such assumptions obviously need to be substantiated by large-scale empirical work and more rigorous methodology. To test its universality, follow-up studies will especially have to examine the specific types of language metonymies used in particular languages and compare them across languages and language families.

Given the limited goal of this study, the paper addresses the following issues: Section 1 looks at the metonymic chain leading from ‘speech organs’ to ‘language.’ For reasons of simplicity, ‘voice’ will be treated jointly with ‘speech organs.’ The linguistic analysis in the ensuing sections will proceed in the order in which the metonymic expressions refer to articulators in the vocal tract, starting from the lower places of articulation. Accordingly, Section 2 considers metonymies and metaphors based on voice, Section 3 investigates metonymies and metaphors related to the tongue, Section 4 explores metonymies and metaphors connected to the teeth, Section 5 looks at metonymies and metaphors dealing with the mouth, Section 6 considers the lips as a metonymic and metaphorical source, and Section 7 summarizes the findings and draws conclusions about the folk model of language.

1. The metonymic chain from ‘speech organ’ to ‘language’

Like the expert model of language, the folk model comprises different levels. The following four levels may be distinguished, which display increasing degrees of complexity or abstractness: (i) articulation, focusing on voice and speech organs such as the tongue, (ii) speaking, including various aspects related to speaking such as gossiping, (iii) speech, i.e. parole, focusing on spoken language, and (iv) language as a system, i.e. langue. These levels of the folk model interact in such a way that one can metonymically shift from a lower level to a higher level, i.e. from the set of speech organs and articulation to speaking, from speaking to speech, and from speech to language, or by skipping one of the intermediate levels, for example by shifting immediately from the level of articulation to that of language.

1.1. speech organ for speaking

Speech organs play a prominent role in the folk model of language: they are instrumental to the action of speaking. The conceptual link between instruments and actions performed with them is well established:[6] instruments may in general stand for actions in which they are used. In English, the metonymy instrument for action is, for instance, found in noun-verb conversions such as to ski or to hammer. Likewise, a speech organ may, as a type of instrument, metonymically stand for the action of speaking. An example of the metonymy speech organ for speaking is Mary mouthed a prayer. Here, the speech organ mouth metonymically stands to the action of saying a prayer. Speech organs of course only serve as instruments for articulation, but in our folk model of language they are also seen as conveying meaning. For example, the expression articulate in Mary is very articulate does not, or not only, refer to her careful enunciation but also, and more importantly, to her skillful and eloquent ability of expressing herself.

1.2. speaking for speech

One’s actions of speaking result in coherent speech. We generally conceive of the results of actions metonymically in terms of the actions bringing them about. The metonymy action for result is, for example, found in nominalizations such as talk from to talk. Its specific application in the domain of language is found in the result nominal speech itself.

1.3. speech for language

Speech is a specific instantiation of language in general, and the metonymic relationship between the two notions of parole and langue might be captured as one of ‘specific’ and ‘generic.’ The metonymy specific for generic is well attested. For example, proverbs usually describe a specific situation but are meant to convey a general understanding. Thus, One swallow does not make a summer is understood to mean more generally that ‘one single piece of evidence is not enough to prove anything.’ In the domain of language, the metonymy speech for language is, for example, attested in the origin of the words for ‘language’ in the Germanic languages (Swedish språk, German Sprache, Dutch taal, etc.) and English speech, as in the speech of lower-class children, in the sense of ‘variety of a language.’

1.4. The metonymic chain

The metonymies discussed form a chain leading from the articulatory function of ‘speech organs’ to articulatory and semantic aspects of ‘speaking’ to the specific result of ‘speech’ and eventually to the generic notion of ‘language.’ This metonymic chain is illustrated in Figure 1.

general metonymies		(i) instrument for action			(ii) action for result			(iii) specific for generic
specific metonymies		speech organ for speaking			speaking for speech			speech for language
	speech organ		>	speaking		>	speech		>	language

The naturalness of the order in this metonymic chain can be seen from the fact that it is much harder, or near-impossible, to think of metonymies displaying the reverse order such as ^?language for speech, ^?speech for speaking or *language for speech organ. The order in the metonymic chain is motivated by cognitive principles governing the selection of preferred metonymic vehicles (see Radden and Kövecses 1999): the physical (or physiological) nature of speech organs outranks the action of speaking, which in its turn outranks the more specific nature of speech, which again outranks the abstract nature of language.

Ideally, we might expect to find metonymies operating step by step along this chain. Many language metonymies, however, skip intermediate stages. Thus, a metonymy widely attested across languages, speech organ for language, links the first and the last elements of the chain as in the metonymic expression tongue for ‘language.’ The ensuing discussion will not go into the details of the intermediate stages.

1.5. The articulatory basis of the folk model of language

Our folk understanding of speaking and language is thoroughly shaped by its physiological, i.e. articulatory, basis. Of the components involved in the production of speech, it is particularly the middle and the upper articulators that the folk model focuses on: the tongue, the mouth and its most visible part, the lips. These speech organs are the most salient articulators: the speaker experiences their changes in position and shape in producing words, and the hearer can see, and pay attention to, these speech organs as they change in speaking. Not surprisingly, therefore, words for the tongue, the mouth and the lips are typically found in the metonymic conceptualization of speaking and language. The perceptually ”hidden” places of articulation of the vocal tract are also conceptually less salient. The larynx, the uvula, the velum, the palate, the alveolar ridge and the teeth are not seen as essential to articulation in folk phonetics, and hence are only exploited very little for metonymic shifts. Voice, however, is recognized as an essential component of articulation in the folk model and is therefore used as a metonymic vehicle. We will start the discussion of language metonymies by looking at voice first.

2. Voice

In the Indo-European languages, words for ‘voice’ mostly derive from, and are connected with, words for ‘speak, say,’ and ‘sound’ (Buck 1949: 1248). It is, of course, an open question whether the folk notion and the linguistic notion of ‘voice’ are the same. Since most sounds are voiced, voice may also be taken as the salient default case and stand for spoken sounds in general. The semantic development of Gr. phÇnJ ‘voice’ to ‘sound’ may support this view. ‘Voice’ is also associated with the place where sounds are produced, the throat, and, as its larger body part area, the neck. Mike Cahill pointed out to me that in Konni, a Gur language spoken in northern Ghana, the word for ‘voice’ is the same as that for ‘throat.’ In Tok Pisin the word nek means both ‘neck’ and ‘voice.’

2.1. voice for speaking

The conceptual relation between voice and speaking is obvious: producing voice is a precondition for articulating sounds and ultimately for speaking. Unlike the spoken sounds of a language, the quality of voice is, however, highly unspecific. Basically, voice can only be ”turned on” or ”turned off” and, if it is on, ”turned up” or ”down.” The metonymic equivalents therefore typically relate to the beginning (1a), the process (1b) or the ending of speaking (1c) or to the volume of speaking (1d):

The metonymy voice for speaking is not very widespread cross-linguistically—languages much rather tend to code these situations non-metonymically by expressions of ‘speaking’ as in the English glosses above or by means of the metonymy word for speaking as in German das Wort ergreifen (the word grasp = ‘to begin to speak’). The metonymic senses ‘voice’ may, however, give rise to various implicatures as in the following examples:

Sentences (2a) and (2b) illustrate implicated extensions of the ‘voice’-metonymy in the domain of turn-taking, and sentences (2c) and (2d) examplify different implicatures invited by the volume of one’s voice in speaking: a person speaking in a loud voice is seen as being opposed to something in Bosnian and as showing off in Italian. Both implicatures are based on the same metonymy: effect for cause, i.e. I raise my voice because I am opposed to something or because I want to show off.

The range of metonymic extensions of the concept ‘voice’ is not very productive and tends to be idiomatic. For example, in Polish the transitive verb g»osiƒ (voice.verb) only applies to a limited set of objects such as theories (g»osiƒ teori voice theory.acc = ‘propose a theory’), and in English to voice is typically used with opinions (to voice an opinion). Voice is used as a functional verb here rather than a content verb, as can be seen from its failure to allow pseudo-clefting: ^?What he voiced was his opinion. All these restrictions in the metonymic usages of ‘voice’ result from the conceptual destitution of the literal concept ‘voice.’

Some metonymic extensions of ‘voice’ are found cross-linguistically: the sense of ‘vote’ is found in, for instance, German Stimme, Polish g»os or Kurdish deng (voice), that of ‘unanimous’ underlies expressions such as French d’une seule voix, German mit einer Stimme and einstimmig or Polish jednog»oÑny as well as English the voice of the people or Italian la voce del popolo. All these metonymic expressions are in conformity with the undifferentiated meaning of voice as an ”on/off” phenomenon: a vote is cast for a party or person, and a unanimous decision or expression is something supported by everybody. The expression ‘voice of the people’ in different languages is a loan translation from Latin vox populi. However, this in itself does not explain its wide-spread use—it has been adopted as a calque in so many languages because the ‘voice’-metonymy is conceptually well motivated.

2.2. voice for language

The notion of ‘voice’ hardly lends itself as a metonymic vehicle to the highly diversified concept of ‘language.’ Yet, there are languages which refer to language by way of ‘neck’ or ‘throat,’ which, as mentioned above, in their turn metonymically stand for ‘voice.’ Ross Clark pointed out to me that the words for ‘voice, language’ in Austronesian languages, e.g. reo in Maori, are derived from words meaning ‘neck.’ Peter Jacobs mentioned as a further example the North-American Indian language Squamish, a coast Salish language, where the suffix –qin ‘hair, throat, head’ is used to refer to names of languages: thus ‘speaking the Teit language’ is rendered as teyt-qin ‘Teit-throat.’

3. The tongue

The tongue is the speech organ that is most conspicuously involved in the action of speaking. The concept of ‘tongue’ is therefore conceptually contiguous to that of ‘speaking’ and forms a well-motivated metonymic vehicle for ‘speaking,’ ‘speech’ and ‘language.’ This is also reflected in visual representations of speech. Ivan Derzhanski pointed out to me that in Aztec pictography, speech was designed by drawing a loose tongue in front of the mouth of the speaker. The concept of ‘tongue’ is also typically used as a metonymy for ‘language.’

3.1. tongue for speaking

In our folk model of language, the tongue enables people to speak, and its pure presence ensures a person’s ability to speak as in:

The conceptual link between having a thing and making use of it is commonly exploited metonymically as precondition for action. Not having the thing required for the action consequently means ‘not being able to perform the action.’ In the case of linguistic action, this situation is typically described by stating the cause for its lack, i.e. by means of the cause for effect metonymy, and by expressing the cause metaphorically. The cause is metaphorized as the loss of one’s tongue (4a), as forgetting one’s tongue (4b), as swallowing one’s tongue (4c) and as having one’s tongue tied (4d), stuck (4e) or glued (4f):

The image of the tongue sticking to the roof of one’s mouth in (4e, f) was already used in Job 29, 10 (”their tongue cleaved to the roof of their mouth”) and probably alludes to the parched tongue of a panting dog hanging out. Irrespective of its Biblical origin, the metaphor is well-motivated as a suitable description of a person’s lack of speech.

In the logic of this folk model, making or allowing a person to speak is seen as loosening his or her tongue (5a) or frenulum (5b):

From the present-day point of view, untying a person’s tongue is, of course, understood metaphorically. Cutting the frenulum of new-born babies was, however, commonly practiced by midwives until recently in order to enable children to learn to speak faster. This is also reflected in expressions such as Norwegian hun er ikke skDret for tungebDndet (she is not cut for the tongue-cord = ‘she has a ready tongue’) or Dutch hij is goed van de tongriem gesneden (he is well of the tongue-cord cut = ‘he is very eloquent’). Thus, these expressions have a literal origin, but the practice itself derives from the firm belief in the folk model of language.

The tongue needs to be unimpeded so that it can move freely to produce speech. Speaking is therefore tantamount to moving one’s tongue (6a), and the kind of motion the tongue typically performs in speaking is turning (6b). The tongue itself needs to be flexible and cannot, for example, be stiffened by having bones (6c):

‘people will talk as they wish,’ i.e. ‘we can’t stop people from talking’

Deliberately refraining from speaking is often metaphorized as in English holding one’s tongue, as bridling one’s tongue (7a), as drawing one’s tongue back (7b) or as putting one’s tongue behind one’s teeth (7c), while silencing a person may be metaphorically described as cutting one’s tongue (7d):

When a speaker decides to refrain from speaking he or she usually does so for some reason so that these situations may give rise to implicated meanings. For example, while (7b) ‘drawing back one’s tongue’ in Bosnian only means ‘to stop speaking,’ the Dutch equivalent zijn tong inslikken is further extended to mean ‘not saying what you want to say’ or ‘breaking one’s promise or word,’ and while (7c) ‘putting the tongue behind the teeth’ in Bosnian means ‘to become silent,’ Dutch zijn tong achter zijn tanden houden means ‘not showing one’s thoughts.’ The more urgent the need for a speaker to stop speaking, the more drastic are the measures to be taken metaphorically in suppressing one’s speech. Such a situation typically arises at the moment when the speaker realizes that he or she is at the point of giving away a secret s/he was not supposed to tell. Many languages describe this situation as biting one’s tongue as e.g. in Polish ugryÑƒ si w jzyk (bite refl in tongue = ‘nearly say something that one shouldn’t say, but could have said’) or even as biting off one’s tongue as in Hungarian inkább leharapná a nyelvét (rather off.bite.3sg.conj def.art tongue.poss.acc = ‘not revealing a secret under any circumstances’). In all these examples, the act of speaking which the speaker stops at the last moment does of course not refer to the form of the message but metonymically refers to its content or meaning.

The folk model of language distinguishes between the form and meaning in language. As is known from Reddy’s (1979) conduit metaphor, linguistic expressions are metaphorized as containers filled with meanings. In the process of speaking, the word containers are placed on the tongue to be filled with meaning and spoken. This process may, however, be blocked so that a word container is seen as lying or revolving on the tongue waiting to be filled with meaning. This metaphor does not, however, describe a situation in which we know a word form but cannot find its meaning but describes the opposite situation: we have a certain concept and cannot find the correct word to express it. In spite of its blatant inappropriateness, this part of the folk model is very widespread cross-linguistically and only differs with respect to the part of the tongue where the word container is placed or whether it is stationary or moves around. Typically, the word is on the tip of one’s tongue as in English or on the end of one’s tongue (8a), on one’s tongue as a whole (8b) or underneath one’s tongue (8c), and it may lie on one’s tongue (8d) or even revolve on it (8e). All these expressions mean ‘not being able to say something although one knows it.’

In the logic of this model, forgetting something is metaphorically seen as a thing falling off one’s tongue as is done in Finnish:

Since the tongue is both responsible for articulation and sense-giving, it may fail with respect to either task. Difficulty in articulation is generally metaphorized as the twisting of the tongue as in English tongue-twister or as breaking of the tongue (10a), failure in communicating the intended meaning is often metaphorized as in English slip of the tongue (10b) or as stumbling of the tongue (10c):

In Japanese, a slip of the tongue, sitaga suberu (tongue.nom slip), refers to the manner of speaking and means ‘saying something carelessly.’ The manner of speaking tends to be associated with the shape and condition of the tongue and its speed of motion, all of which provide a rich source for metaphorization. The length of the tongue correlates with the length of speaking, which in its turn gives rise to various implicated meanings: gossiping in Dutch (11a), blabbing and giving away secrets in Polish (11b) and Italian (11c), and being insolent and impertinent in Kurdish (11d) and Turkish (11e):

The metaphorical use of a ‘short tongue’ was not attested in the languages studied. However, the tongue may be both strong (Bosnian jak ‘confident’) and weak (Bosnian slab ‘poor in speaking’), it may be sharp (i.e. ‘critical’) and sweet (i.e. ‘flattering’ as in English, or ‘polite, friendly’ as in Kurdish zimanÕTrTn ‘tongue sweet’); it may be good or clean (Dutch goed, Kurdish zimanzelal tongue-clean = ‘eloquent’) and bad (Bosnian biti pogana jezika be bad.gen tongue.gen or imiti pogan jezik have bad.acc tongue.acc = ‘be slanderous’); it may be big (Bosnian velik(i) jezik big tongue = ‘be bragging’), fast (Dutch rad ‘able to speak well’), etc. Most of these metaphorical senses are transparent and need not be analyzed.

The associative link between a ‘good tongue’ and ‘eloquence’ is particularly revealing for our folk model of language. It is also reflected in the derivation of words for eloquence. Thus, Arabic lasina ‘to be eloquent’ and lasan ‘eloquence’ derive from lis~n ‘tongue,’ and Japanese benzetsu ‘eloquence’ is composed of ben ‘valve, speak’ and zetsu ‘tongue.’ Thus, making use of one’s tongue properly means being eloquent. Speaking properly also means speaking sincerely. The tongue should, for example, be held in the right position (12a, b) and not be misplaced (12c), be split (12d) or come in the shape of two tongues (12f):

As was shown in the examples under (3), the presence of the tongue is needed for speaking. But in order to say something meaningful, the tongue needs to be monitored by the mind. If this is not the case, the tongue will produce nonsensical utterances (13a, b).

The tongue is seen as the speech organ which gives words their meanings. The tongue expresses our sincere thoughts (14a) and vents our true feelings (14b, c), i.e. it makes us say what we mean:

The folk model does not restrict the gift of speaking to humans, but also allows objects to speak as in Money talks. Also the semantic relationship between Polish rzec ‘say’ and rzecz ‘thing, language’ appears to be based on this folk view. In Arabic, situations in which animates or things seem to say something are expressed by lis~n ‘tongue, language.’ With inanimates the expression lis~n al_hal (tongue def.situation) refers to a state of reality which may be paraphrased as ‘as reflected by,’ with animates, the expression lis~n hal-ih-i (tongue situation.anim.poss) is understood in the sense of ‘as if he wanted to say ...’ or ‘his present state is indicative of ...,’ as in the following sentence:

The use of lis~n among the three words for ‘language’ in Arabic—the other two, lugha and kalaam, being derivatives from words for ‘speak’—is revealing: things as well as physiological expressions cannot but communicate their true nature. Metaphorizing these situations as ‘speaking with a tongue’ is, therefore, the most appropriate choice.

3.2. tongue for language

Being the principal articulator in the production of speech, the tongue is also the most appropriate metonymic vehicle among the speech organs to be selected for the more abstract notion of ‘language.’ The lexicalized metonymy the tongue for language is especially prevalent in the Indo-European languages: Greek glôssa, Latin lingua and Romance languages, Common Germanic *tuõgÇn and Germanic languages, Slavonic languages (Church Slavonic jzykß, Polish jzyk, Bosnian jesik, Bulgarian ezik, Russian jayzk), Baltic languages (Lithuanian liežuvis, Latvian mle), Indo-Iranian languages (Urdu and Persian zabaan, Kurdish ziman), etc. In fact, there does not seem to be any Indo-European language that does not use ‘tongue’ for ‘language.’ Among the non-Indo-European languages, the word for ‘tongue’ is used for ‘language’ in, amongst others, the Uralic languages (Finnish kieli, Hungarian nyelv), Turkic languages (Turkish dil), Semitic languages (Hebrew lashon, Biblical Aramaic lishan, Arabic lis~n), Caucasian languages (Georgian ena, Chechen mott), Chadic languages (Hausa harshèe) and American Indian languages (Salish, lexical suffix of shape /*-tsan/).

The tongue for language-metonymy is widespread across languages, but is of course not universal. In some languages, this metonymy competes with the metonymic use of other speech organs or, more frequently, with the notions of ‘speaking’ or ‘speech’ for ‘language.’ In this case, the word for tongue may be restricted to specific varieties of language. For example, one’s native language is typically described as ‘mother tongue,’ and a regional variety is often referred to as ‘local tongue.’ Finally, the Biblical use of the term tongue in ‘speaking in tongues’ to describe divine, but unintelligible, speech may have contributed to associating the tongue with very specific varieties of language.[7]

4. The teeth

The teeth contribute to articulation, but the folk model of language does not assign any major articulatory function to the teeth. Some languages have fixed expressions contrasting the teeth with the tongue or the mouth: the tongue enables us to speak, whereas the teeth lack the faculty of speaking. If a person wanted to speak with his or her teeth instead of with his or her tongue, s/he is silent (16a) and, in metonymic reference to meaning, does not know what to say (16b):

Most importantly, the teeth need be opened in speaking as indicated by the Chinese expression qi-chi (open-teeth ‘start to talk about something’). In Chinese, the pointed shape of the teeth allows the speaker to hang ideas on them: ‘to mention,’ i.e. to talk about something, is metaphorically expressed as gua-chi (hang on-teeth). The teeth do not interfere with the meanings of the words generated by the tongue. They are even associated with frankness, as illustrated by the Bosnian adjective zubat (tooth.adj), which is derived from ‘tooth’ and is used to describe a person who openly says what he or she means. Likewise, speaking with one’s teeth conveys frankness as in Bosnian (17a), and this may be the reason why the teeth should not be ”covered” as in Japanese (17b):

The teeth may, however, obstruct the free passage of the word on its way from the tongue to the mouth. Speaking through one’s teeth leads to low speech (18a) or murmuring (18b) and, by implicature, to reluctance in answering (18c, d):

The teeth may also be opposed to the tongue. This is reflected in the Bulgarian saying da sa zhivi predni zrbi (long live the front teeth), which Ivan Derzhanski mentioned to me. The saying is used when one narrowly avoids saying something secret, tactless or otherwise inappropriate. This expression could be translated as ‘bless my front teeth’ and conveys the idea that the tongue, which generates speech, can be too zealous sometimes, so it is the teeth’s duty to stop it.

In all the examples listed the teeth are not seen as producing speech on their own but only as modulating speech or preventing it from being uttered. There is thus no basis for the metonymy *teeth for speaking or *teeth for language.

5. The mouth

In the folk model of language, the mouth as a whole figures as a speech organ, but it may also metonymically stand for one of its articulatory regions as its ”active zone”[8], in particular the lips. The importance of the mouth for speaking can be seen from Chinese and Japanese ideograms for linguistic activities and language. Thus, the Japanese kanji character for ‘say,’ gen, iu, shown in Fig. 2a, is composed of the idiogram of a square for ‘mouth’ and an ideogram of three horizontal lines and a slanting stroke above it, which originally represented a needle and the idea of ‘sharp.’ This ideogram might indicate articulate use of the mouth or the needle as depicting teeth, which were considered necessary for good enunciation (Henshall 1998: 81). The sign for ‘speech, talk,’ hanasu, wa, shown in Fig. 2b, is composed of the above ideogram for ‘say’ and an ideogram originally representing a hollowed-out space, which lends its sound to express ‘good,’ thus ‘good words,’ giving ‘speech’ (Henshall 1998: 63). Interestingly, the second ideogram is usually, but incorrectly, identified as a tongue, giving the meanings of ‘articulated words’ and hence ‘speak.’

5.1. mouth for speaking

The primary and phylogenetically prior function of the mouth is, of course, that of taking in and chewing food. Surprisingly, however, hardly any elements of the food domain are metaphorically carried over onto that of speaking—the target domain typically understood in terms of food is that of ideas. Expressions of ‘speaking’ that relate to the domain of eating are mouthful (amount of food put in one’s mouth = ‘comment or remark rich in meaning’) and Hungarian a szájába rág valakinek valamit (art.def mouth.poss.in.direction chew.3sg someone.dat something.acc = ‘to explain to someone at full length what to say’). The relationship between ‘mouth’ and ‘speaking’ is predominantly metonymic: the mouth is regarded as a speech organ that is involved in speaking. This view is reflected in the Japanese word for ‘speak,’ kutio kiku, which literally translates as ‘mouth.acc be.effective,’ or the Bosnian word for ‘say,’ izustiti, literally ‘out.mouth.inf.’ An interesting piece of indirect evidence for the mouth for speaking-metonymy has been provided to me by Marika Butskhrikidze. The Georgian expression for ‘domestic animal,’ p’irut’qvi, is composed of p’ir- ‘mouth,’ u- negative marker, t’q’v- a root from infinitival met’q’veleba ‘speaking,’ and the nominative case marker –i. The word thus defines the domestic animal by its property of lacking a mouth for speaking. Conversely, we may assume that one of man’s outstanding characteristic is having a mouth for speaking.

The mouth and the tongue may be seen as jointly contributing to speaking, as can be inferred from Chinese expressions in which both articulators are responsible for abnormal speech. Thus, using awkward, slow or inarticulate speech is described in Chinese as ben-kou zhuo-she (awkward-mouth clumsy-tongue), being tongue-tied or at a loss for words as zhang-kou jie-she (open-mouth knotted/stiffened-tongue), and talking in a tart and mean manner as jian-uzui bo-she (sharp/pointed-mouth thin-tongue). The mouth is, however, only of secondary importance to the tongue as suggested by the Turkish and Arabic sayings in (19a, b), which refer to a person who is helpless or too shy to speak:

The mouth is less mobile and flexible than the tongue; moreover, the mouth as a whole contains several articulators whose specific function it may take over. For example, by opening and closing one’s mouth, the speaker may begin and stop speaking in the same way that he may turn his voice on or off. Thus, opening one’s mouth only means ‘to start speaking’ in (20a), but may also give rise to implicatures as in (20b), and closing or holding one’s mouth is almost exclusively used as a rude request to stop another person from talking (20c, d), while deliberate closing of one’s mouth refers to the speaker’s intention of keeping his or her thoughts to him- or herself (20e):

Shutting one’s mouth may be metaphorically achieved in different ways. Not saying a word is metaphorized in Bosnian as having the mouth tied up (usta mu se zavezala ‘mouth him refl tied’) and in Italian and French as having the mouth sewn up (avere la bocca cucita ‘have the mouth sewn.up’). Possible ways of preventing the contents of a container from leaking or getting out is to plug, lock or seal it—hence, in order to silence a person, the mouth as a hollow space may be plugged as in Italian (tappare la bocca a qualcuno ‘plug the mouth of someone’), locked as in Swedish (sätta munlDs pD nDgon ‘set lock on something’) or sealed as in English to seal someone’s lips.

A number of linguistic functions are uniquely associated with the mouth as a container. The interior of the mouth may be filled with the form of a single word that is difficult to pronounce (21a) or a language as a whole (21b), the speech in one’s mouth can be seen as a treasure (21c), and a person whose mouth is filled is boastful (21d):

Typically, the words to be spoken are in the speaker’s mouth, and many things may happen to them before they can be uttered. One may, in particular, put words into another person’s mouth and thereby tell the person what to say (22a), take words out of a person’s mouth and thereby say the same thing that the speaker was going to say (22b), or turn words around in the speaker’s mouth and thereby distort the meaning of what he or she said (22c):

The meanings these metaphors convey are consistent with the folk model of language: Since the speaker decides on the meaning to be given each word, speaking another person’s words as in (22a) runs counter to his or her own intentions. By A’s taking a word from B’s mouth as in (22b), B no longer has the word so that A, but not B, can now utter B’s word. By turning a word around in a person’s mouth as in (22c), its proper position is affected and, concomitantly, its meaning. The metaphor is, however, inconsistent in that the word is no longer in the mouth but has already been uttered.

Like the tongue, the mouth may be associated with certain properties which metaphorically describe aspects of speaking. While the set of properties which are attributed to the mouth is different from that of the tongue, the meanings metaphorically conveyed are often similar to the ones typically expressed by the tongue. Thus, a healthy mouth stands for eloquence as in Japanese kutiga tassha da (mouth.nom healthy be), while a sick mouth stands for offensive words as in Hausa ciiwòn bàakii (illness.of mouth), a light mouth characterizes someone who reveals secrets easily as in Japanese kutiga karui (mouth.nom light), and, conversely, a hard mouth characterizes a person who does not reveal secrets easily as in Japanese kutiga katai (mouth.nom hard). A person who feels self-important has a big mouth as in Bosnian imati velika usta (have big mouth = ‘to be self-important’), while someone who is talkative or quarrelsome has a wide mouth as in Polish mieƒ gb od ucha do ucha (have mouth.acc from ear to ear = ‘to be talkative, quarrelsome’). People may also have loose, pleasant or golden mouths, which are associated with fairly obvious personal traits.

Like the tongue, the mouth should have the proper shape and occupy the proper position when speaking. A person who leaves his mouth open (23a), speaks with two mouths (23b), speaks with half a mouth (23c) or even speaks without his or her mouth (23d, e), or a person who plays with his or her mouth (23f) does not communicate properly:

It follows from the above discussion that people who speak ”with one mouth” are in agreement (24a), and if their mouths are different as in (24b), they disagree with one another:

Like the tongue, the mouth may speak independently of the heart as revealed by the French old-fashioned saying under (25a). The mouth may also be in disagreement with other speech organs, in particular the teeth as in the Hausa saying under (25b), which more generally refers to friction which is bound to occur between friends unless there is some ‘give’ and ‘take’ between them as in (25c):

3ms.compl come.out mouth and teeth 3ms.rel.compl say ‘he came out with it’ = ‘he then openly said ...’

5.2. mouth for language

The word for the mouth as a whole is occasionally used as a metonymic vehicle to refer to ‘language.’ It is typically used for oral, as opposed to written, varieties and regional varieties. Thus, the German and Swedish words for ‘dialect,’ Mundart and munart, literally mean ‘mouth-manner,’ i.e. ‘manner of speaking.’ The Gaelic word for ‘English,’ Beurla, derives from beul ‘mouth.’ Ivan Derzhanski informed me that the Quechua word for its language is runa simi (man mouth, i.e. ‘human language’). Ron Cosper drew my attention to Chadic languages, where the word used for ‘language’ means ‘mouth,’ e.g. Polci bii, Sayanci vii, Boghom pyok, all of which are cognate with Hausa bàakii ‘mouth.’ Peter Jacobs pointed out that in the North-American Indian language Squamish and in probably all the other languages in the Salish language family, the suffix indicating mouth is used to refer to language and speech. Alexandra Aikhenvald provided the example of Tariana, a South American language of the Arawak family in Northwest Amazonia, in which the word for mouth, -numa, has become grammaticalized as a classifier for ‘word.’

6. The lips

The lips are the visible part of the mouth and form the upper end of the vocal tract. In some languages, the word for lip and mouth are related or the same. For example, Bosnian usna ‘lip’ derives from usta ‘mouth’ via ustna ‘mouth.of,’ and in Hausa the lips are referred to as ‘skin of mouth’ (faatà+ bàkà). Hence, many of the articulatory functions attributed to the mouth also apply to the lips. The lips play an important role in our folk model of language as can be seen from the widely overestimated belief in lip-reading, i.e. expert’s and deaf person’s ability of understanding a speaker’s words and their meanings from watching the movements of his lips. The lips are, however, only the final articulators to shape the quality of the sound produced. Their contribution to articulation is, therefore, smaller and more restricted than that of the tongue.

6.1. lips for speaking

The role the folk model of language assigns to the lips in an act of speaking is mainly that of opening or closing the mouth and thereby allowing words which have been produced by the tongue to be uttered or not to be uttered. Thus, ‘to utter’ and ‘utterance’ in Kurdish is rendered as lLv-kirin (‘lip do’). The default situation of speaking tends to be seen as having the mouth open rather than the lips—open lips may serve to characterize a person as in Dutch loslippig ‘talkative, prone to giving away secrets.’

Deliberately keeping one’s lips closed tends to be associated with ‘keeping a secret’ as in Dutch de lippen op elkaar houden ‘the lips on one another hold’ or English my lips are sealed. We also see the speaker closing his or her lips in expressions which literally refer to words not coming over one’s lips as in Norwegian Ikke et ord kom over hans lepper ‘not a word came over his lips’ or in the commonly used metaphor of biting one’s lips as in Polish przygryïƒ wargi ‘bite lips.’ Since the lips are the last speech organs to prevent a word from slipping out, the lip-biting image appropriately captures the emergency situation of suppressing an embarrassing remark at the last moment. The reason for not speaking may, however, also be the speaker’s word-finding problem (26a) or his or her decision to stop speaking (26b):

Having a word on one’s lips as in (26a) is a metaphorical variant of having a word on one’s tongue as in the examples under (8), and in the same way that the word container may more specifically be placed on the tip of one’s tongue, it may be placed at the edge of one’s lips as in French Je l’ai sur le bord des lèvres (I it have on the edge of.pl lip.pl), implying that the speaker is so close to finding the word he or she is thinking of that it is almost uttered. This image as well as that of a word ”dying on one’s lips” in (26b) reflects an underlying folk view in which words pass through the vocal tract from the tongue to the lips, where they are ejected in speech or, in these cases of becoming silent, stop at the lips.

The lips’ final impact on a word’s articulation seems to be minimal and, by metonymy, the lips do not affect a word’s meaning. By attending to the speaker’s lip movements, a hearer can therefore ”read” their meaning or, if he or she wants to listen more attentively to an interesting topic, ”cling” to the speaker’s lips (27a) or ”hang” on his or her lips (27b). Both of these metaphorical expressions are of course based on the conceptual metaphor seeing is touching.

Like the tongue, the lips may thus truly express one’s thoughts as in Italian avere il cuore sulla labra (have the heart on the lips = ‘to be sincere’) or Hungarian ajakima (lip.prayer = ‘prayer expressed by words’). More commonly, however, the lips are seen as changing a word’s original meaning and, thereby, distorting it. In this case, the speaker only pays ”lip service,” confesses from the end of his or her lips (28a) or prays by using his or her lips only (28b). When, as pointed out to me by Joe McIntyre, a Hausa speaker makes empty promises, s/he uses his or her lips as in (28c) or metaphorically speaks of the lips of male camels as in (28d):

It is probably for this reason that expressions for speaking involving one’s lips tend to carry negative connotations. For example, in Japanese phrases containing the word for ‘lip,’ kutibiru, denote negative aspects of speaking: to be talkative (kutibiruga usui lip.nom thin), to be critical (kutibiruo hirugaesu lip.acc turn-over), to be hateful (kutibiruo kaesu lip.acc return, turn over), and to complain (kutibiruo togarasu lip.acc sharpen). The English threat Watch your lip! and the German colloquialism eine Lippe riskieren (a lip risk = ‘to be cheeky’) also have aggressive undertones. The same applies to Chinese, where the conjunction of ‘lip’ and ‘tongue’ leads to the sense of ‘argument’ as in chun-she (lip-tongue ‘words, argument’) and chun-qiang she-jian (lip-spear/gun tongue-sword ‘cross verbal words, engage in a battle of words’).

6.2. lip for language

Very few languages seem to use the word for lip in the sense of ‘language.’ Examples include Irish barla, derived from bl ‘lip’ and used especially in reference to the English language, and Israeli Hebrew safa ‘lip, language.’ Safa is generally assumed to be synonymous with another body-part term for ‘language,’ lashon ‘tongue.’ According to the different roles assigned to the tongue and the lips in our folk model of language, we should, however, expect to find differences in meaning between safa and lashon. We will conclude this section by comparing the ranges of application of these two terms for ‘language’ as provided to me by Bill Morris and Yehuda Falk.

Safa is the more common term and applies to language in general. It is used in expressions denoting natural language (safa tiv’it ‘language natural’), spoken language (sfat dibur), sign language (sfat simanim), artificial language (safa melaxutit), programming language (sfat tixnut), foreign language (safa zara ‘language foreign’), people who speak foreign languages (dovrei safot ‘speakers.of ‘safa’.pl’) and the ‘War of languages’ (hasafot) before World War 1, when German was about to take over and Hebrew was desparately striking back. The collocations in which safa is found reflect the wide range of pragmatic purposes and functions in which speech is used in everyday situations: as plain language and clear speech as well as incomprehensible, verbose, long-winded and flowery language, as conveying words of truth and promise as well as sweet words, flattery and lies, as slandering, jabbering, chattering, bragging, stammering, speaking harshly, half-heartedly, reluctantly or hypocritically.

Lashon, on the other hand, is restricted to traditional, religious language. It is used in expressions for The Holy Tongue (lashon hakodesh), Biblical Hebrew (leshon hamikra) as well as other periods of Hebrew, Mishnaic Hebrew (Mishna is the oldest authoritative collection of Jewish oral law), foreign language (Lashon Am Zar or L-A-Z ‘language of anyone not Jewish’), scientific language, linguistics and linguist (balshan/ut), and is used in the name of the Academy of Hebrew language. Lashon has, according to one informant, a Shakespearean or Biblical feel to it. The literary and Biblical connotations associated with lashon are reflected in collocations meaning ‘fine language,’ ‘exactly as written,’ ‘be precise in one’s speech’ and ‘understatement.’ Lashon is, however, also used in collocations for language in general as in ‘everyday speech’ and ‘common parlance,’ which are typically expressed by safa, and with negative connotations as in ‘flattery,’ ‘smooth talking,’ ‘gossip,’ ‘scandal’ and ‘slander.’

Although a few aspects of meaning are shared by safa and lashon such as ‘flattery’ and ‘slandering,’ their prototypical senses are different. Thus, lashon is the appropriate word to use in a sentence such as Ivrit zot lashon hakodesh ‘Hebrew is the Holy Language,’ but not, for example, in ^?eize leshonot atah medaber? ‘which language do you speak?,’ where safa would have to be used. The ”tongue word” lashon applies to a highly respected and meaningful language, which therefore also serves as a means of identification, whereas the ”lip word” safa applies to language as an everyday means of communication with all its shortcomings of misuses.

7. Conclusions

The object of this study was to investigate the folk model of language. Recent cognitive research on metonymy and metaphor has been able to reveal the conceptual underpinnings of many complex and abstract notions. The complex notions which this study was concerned with are those of ‘speaking’ and ‘language.’ A particularly salient conceptual source for these notions is their articulatory basis, on which this study has concentrated. There is ample cross-linguistic evidence for a metonymic link between the articulatory notions of ‘voice’ and ‘speech organs’ and the notions of ‘speaking’ and ‘language,’ which was more generally described as speech organ for speaking and speech organ for language. The specific roles assigned to the speech organs in these metonymies reflect a folk model of language.

The articulatory level of the folk model of language includes voice and four articulators: tongue, teeth, mouth and lips. Although the notion of the vocal tract is not explicitly present as a term in the folk model, it figures prominently in the production of speech. Speech production involves the following stages of articulation, each of which may have its own metonymic impact on the meaning of the word articulated.

First, voice is produced in the throat and may be directly meaningful in expressions such as to voice an opinion.

Next, language is generated by the tongue. The tongue is the most important articulator. The folk model of language regards the tongue as the speech organ that not only produces the sound shape of words but also fills them with meaning. The meanings given to words reflect the speaker’s thoughts; hence, in using his or her tongue, the speaker cannot but speak truly and sincerely. Its essential function in producing sounds and generating meaning also makes the tongue the most suitable speech organ to stand for the abstract notion of language.

As a next step, the words are passed on to the mouth. The Turkish expression dil aIz vermemek (tongue.nom mouth.nom not.give = ‘to be silent, not wanting to speak’) refers to this part of the folk model. The mouth is seen as the most important articulator next to the tongue and also contributes to a word’s meaning. On their passage through the mouth the words may, however, be obstructed by the teeth. This idea is reflected in the Texan expression She’s got tongue enough for ten rows of teeth (‘that woman can talk’).

The lips, lastly, are the final articulators to shape the sound and meaning of a word before it is uttered. If they allow a word to pass freely, the hearer can read its meaning from the lips; they may, however, also interfere and distort a word’s meaning before it is uttered.

The discussion of the roles of the speech organs has shown that the folk model of language tends to assign different functions to the speech organs and associate them with different aspects of meaning. In general, the lower, partly hidden articulators in the vocal tract are associated with a ”deeper” quality than the upper, visible articulators. This folk view of language may be related to our metaphorical view of the mind. As shown by Jäkel (1995: 209) in his metaphorical problem-solving scenario, difficult problems lie deep down and are hidden inside the ”problem container” as in There is a deeper issue here. This quality of a difficult problem also requires ”depth of thinking” as in the expression deep thinker in comparison with superficial thinker. The conceptual metaphor involved is intellectual quality is depth of thinking. In the case of lower and upper articulators and the kind of language produced by them, the levels of depth are literally present and the qualities attached to them are invoked by metonymy.

Evidence for this folk model of language has been found in many, also unrelated, languages, and it may even be the case that the basic metonymic structure of this folk model is universal. The metonymies tend to invite conventionalized implicatures, especially in reference to the manner or mode of speaking. These implicatures are pragmatically motivated and are, of course, not universal. For example, we showed that opening one’s mouth may mean ‘to begin to speak’ or ‘to begin scolding someone’ or ‘to say one’s opinion frankly.’ Lastly, the language metonymies tend to be elaborated by metaphors which, in being based on underlying conceptual metaphors, are to a large extent motivated, but not predictable. Thus, the image of describing a word as dying on the lips as in Italian La parola gli morì sulle labbra (the.fem word him die.past on.the lip.pl) is also found in Hausa (yànzu bàakii yaa mutù now mouth 3ms.compl die), but here it does not refer to one’s voice trailing off, but to words failing.

References

Buck, Carl Darling (1949). A Dictionary of Selected Synonyms in the Principal Indo-European Languages: A Contribution to the History of Ideas. Chicago, Illinois.

Goossens, Louis (1995). ”Metaphtonymy: The interaction of metaphor and metonymy in figurative expressions for linguistic action.” In: Louis Goossens et al., eds., By Word of Mouth: Metaphor, Metonymy and Linguistic Action in a Cognitive Perspective. Amsterdam/Philadelphia, pp. 159-174.

Grady, Joseph (1997). Foundations of Meaning: Primary Metaphors and Primary Scenes. Doctoral Dissertation, U.C. Berkeley.

Henshall, Kenneth G. (1998). A Guide to Remembering Japanese Characters. Rutland, Vermont and Tokyo.

Jäkel, Olaf (1995). ”The metaphorical concept of mind: ‘Mental activity is manipulation’.” In: John R. Taylor and Robert E. MacLaury (eds.), Language and the Cognitive Construal of the World. Berlin, New York, pp. 197–229.

Langacker, Ronald W. (1991). Concept, Image, and Symbol: The Cognitive Basis of Grammar. Berlin, New York.

Quinn, Naomi and Dorothy Holland (1987). ”Culture and cognition.” In: Dorothy Holland and Naomi Quinn (eds.), Cultural Models in Language and Thought. Cambridge, pp. 3-40.

Radden, Günter and Zoltán Kövecses (1999). ”Towards a theory of metonymy.” In: K.-U. Panther and G. Radden (eds.), Metonymy in Language and Thought. Amsterdam/ Philadelphia, pp. 17-59.

Reddy, Michael J. (1979). ”The conduit metaphor – A case of frame conflict in our language about language.” In: A. Ortony, ed., Metaphor and Thought. Cambridge et al., pp. 284-324.

[1] According to Gary Palmer (p.c.), the Tagalog word, wikà is exclusively reserved for ‘language’ or ‘language of the people.’

[2] See Buck (1949: 1259-1261) for the origins of words for ‘language’ in the Indo-European languages. The use of expressions meaning ‘speak’ for ‘language’ seems to be even more widespread in non-Indo-European languages. Informants mentioned to me, amongst others, South American languages from the Arawak family, in which the word for language is ‘our talk’ or ‘our speech’ (Alexandra Aikhenvald), the northern Ghana Gur language Konni, where the word for language, balika, is derived from the verb bali ‘say, tell’ (Mike Cahill), or the eastern varieties of Basque, where the word for language, mintzaira, is composed of the verb-stem mintza- ‘speak’ plus the suffix –era ‘way, manner,’ hence, ‘way of speaking’ (Larry Trask).

[3] I am indebted to Larry Trask for this observation.

[4] In Quinn & Holland’s (1987: 4) definition, cultural models ”are presupposed, taken-for-granted models of the world that are widely shared (although not necessarily to the exclusion of other, alternative models) by the members of a society and that play an enormous role in their understanding of that world and their behavior in it.”

[5] Many informants kindly provided me with a wealth of linguistic data from their native tongues, only part of which, however, could be made use of within the limitations of this paper. I would like to express my sincere thanks to Alexandra Y. Aikhenvald, Rita Brdar-Szabó, Marika Butskhrikidze, Mike Cahill, Catherine Chauvin, Ross Clark, Ron Cosper, Claudio Di Meola, René Dirven, Ivan A. Derzhanski, Suzette Haden Elgin, Yehuda N. Falk, Elóbieta Górska, Eitan Grossman, Zeki Hamawand, Peter Jacobs, Karol Janicki, Katalin Jobbágy, James Kirchner, Wojtek Klemm, Mark A. Mandel, Joe McIntyre, Naida Mehmedbegovic, Bill Morris, Thapelo Otlogetswe, Gary Palmer, Aila Radden, Donald F. Reidl, Mari-Ann Richter, Esra Sandikcioglu, Kazuko Shinohara, Jess Tauber, Dominique Thun, Larry Trask, Mahendra K. Verma, Andrew Wilcox and Ning Yu.

[6] The natural link between instruments and actions has already been observed by Plato in his Cratylus. Here, Socrates develops Cratylus’ claim that a name is naturally suited to its object by using the delightful, but faulty, analogy of the actions of cutting, weaving and drilling, whose nature intrinsically requires the employment of certain, natural instruments: a knife, a borer and a shuttle. Likewise, his dialectic argument goes, the action of speaking requires naming as an ”instrument” and, in the same way that a skilled artisan uses instruments, speaking requires a lawgiver to name the objects.

[7] There are several references to the gift of ”speaking in tongues” in the Bible. In Acts 2:4, at the day of Pentecost, the apostles ”were all filled with the Holy Ghost, and began to speak with other tongues, as the Spirit gave them utterance.” Glossolalia is, however, not unique to Christianity but has, according to the Encyclopedia Britannica, s.v. tongues, gift of, ”occurred in some of the ancient Greek religions and in various primitive religions.”

[8] For the notion of active zone see Langacker (1991, Ch. 7 ”Active zones”).

[9] The Swedish word mål and its equivalents in Danish and Norwegian are archaic words for ‘language.’ They derive from ON m~l ‘speech’ and mæla ‘speak’ and are ultimately related to words meaning ‘assembly’ (cf. OE mæzel) and ‘meet’ (cf. OE mtan), i.e. to the place or occasion when speech was used for official purposes. Mål is now restricted to idiomatic expressions such as modersmål (mother.gen.language = ‘mother tongue’) and tungomål (tongue-language = ‘glossolalia’).

The folk model of language

Abstract