Jane and Molly were college students in their early twenties. As Jane was talking on the telephone with her boyfriend Fred, he heard Molly say in the background:
Molly: I’m gonna go over there.
which led to the following exchange:
Fred: Where’s Molly going?
Jane: She’s going next door, to Steve’s?
He’s having some people over.
Fred: Oh.
At which point Jane said:
Jane: Yeah we’re gonna go hang out there.
This chapter traces Jane’s last sentence from its beginning in her thought to the sound she uttered. We can look in turn at its semantic, syntactic, and phonological manifestations, at the processes responsible for each of them, and at the reasons behind those processes.
A book is confined to visual representations, but neither Jane’s thought nor her sound were anything you could see. At best we can only attempt to describe certain aspects of her thought, while representing her sound in one or more of the various ways that are available for that purpose, including the orthographic representation above and any of various acoustic representations such as the display of fundamental frequency at the top of Figure 9.1.
Figure 9.1 Tracking the fundamental frequency of Jane’s sound
Jane’s first word, yeah, was a “discourse marker” (Schiffrin Reference Schiffrin1987), which in this case did not occupy a separate focus of consciousness but was included with the rest of Jane’s utterance within a single intonation unit. Yeah typically functions as a positive response to a yes–no question. In this case the question had not been explicitly asked, but Jane evidently assumed that Fred was wondering whether she and Molly were both going to Steve’s and not just Molly. The next word, we, expressed an idea that was given, but it was spoken with a high pitch because it contrasted Jane and Molly as a pair with Molly alone. Contrastive ideas are expressed with prominent prosody even when they are given.
There is no accepted way to picture a thought, but Figure 9.2 shows the three ideas that were active in Jane’s thought at this moment. The event was the idea of hanging out at Steve’s, its location was Steve’s apartment, and the participants in it were Jane and Molly. The curly brackets are a way of locating these three ideas in the realm of thought. The smaller size of the location and participant ideas shows their status as given, as opposed to the new status of the event. These three ideas were of course not independent of each other but combined to form the totality of Jane’s thought. The event of hanging out was the nucleus of that thought, but it called for the accompaniment of a location and participants.
Figure 9.2 The ideas included in Jane’s thought
A successful conversation requires not only that speakers verbalize their own thoughts but also that they imagine the thoughts of their interlocutors, integrating what they themselves are thinking with what they imagine those others to be thinking. Conversationalists vary in the extent to which they fulfill this requirement, occupying a position on a scale from maximum empathy to maximum solipsism. The difference can appear in the choice of topics and the ways topics are elaborated, but also in the use of pronouns, which ideally express ideas of entities that are assumed to be already active in the other person’s thought, thus establishing them as given. That assumption depends on the prior activation of the ideas in question, as when the idea of Jane was assumed to be active in Fred’s thought because she was his conversational partner, the idea of Molly because of her presence in Jane’s room, and the idea of Steve’s apartment because of Jane’s earlier mention of it.
There were several features that created a distance between Jane’s thought and what would have been directly conveyed by her words if they were taken literally. The words hang and out in themselves bore no direct relation to what Jane was thinking, which included neither hanging nor being out in any literal sense. The word gonna is of course a way of expressing a future intention, with its origin in the sequence going to, but the literal meaning of those words also bore no direct relation to Jane’s thought. On a subtler level the word go in the sequence go hang out was more closely related to her thought – she was indeed planning to go somewhere – but in this case the go expressed an “andative” orientation of the hanging out, and not the separate event it would have been if she had said we’re gonna go there. The example thus illustrates in several ways a separation of Jane’s words, interpreted literally, from what she was thinking. These words, furthermore, were not the only options available to her. With roughly the same thought she might have said, among other possibilities:
Each such alternative, with others that can be easily imagined, would have packaged Jane’s thought in a different way while retaining its essential nature.
From a Thought to a Semantic Structure
We can begin tracing the path from Jane’s thought to her sound by exploring the way her ideas were organized into a semantic structure (Figure 9.3). Once those choices were made, the thought was committed to a particular path of verbalization whose representation in syntax and then in phonology was for the most part already determined. Semantic choices, in other words, are largely responsible for giving language its shape. These choices include selection, categorization, orientation, and combination, each of which we can examine in turn.
Figure 9.3 From a thought to a semantic structure
Selection
A thought probably always contains more than is, or perhaps even can be represented in language. We don’t say everything we think. An essential first step in verbalization is thus to choose how much and what parts of a thought will proceed to be represented with sound. On a large scale that choice can be a commitment to an over-arching topic. In this case there had already been discussion of Molly’s relation to Steve, a topic that was now “in the air.” Speakers navigate through a topic with a succession of thoughts that are often guided by interaction with an interlocutor. Figure 9.4 shows this selection process as the first step toward a semantic structure.
Figure 9.4 Selecting parts of a thought to be verbalized
Categorization
The anticipated event of hanging out at Steve’s was going to be particular in time and space and its particularity called for the next step. It would obviously be impossible for every particular idea to be associated with a particular sound. Franz Boas recognized this problem a century ago: “Since the total range of personal experience which language serves to express is infinitely varied, and its whole scope must be expressed by a limited number of phonetic groups, it is obvious that an extended classification of experiences must underlie all articulate speech” (Boas Reference Boas1963 [1911]: 18).
Not only is the number of thinkable ideas vast if not infinite, but even if the assignment of a particular idea to a particular sound were less problematic than it is, there would be no way a listener could know which idea was associated with which sound. Thought and language solve these problems by interpreting particular ideas as instances of categories. A category serves two functions. On the one hand it triggers expectations regarding an idea, expectations that are typically associated with instances of that category. Perhaps we can lift it, perhaps we can eat it, perhaps we can sit on it, perhaps it barks and chases squirrels. But at the same time a category provides each of its instances with a sound that can be used to verbalize it.
Edward Sapir emphasized the fact that a category can apply to an unlimited range of particular instances:
We must cut to the bone of things, we must more or less arbitrarily throw whole masses of experience together as similar enough to warrant their being looked upon – mistakenly, but conveniently – as identical. This house and that house and thousands of other phenomena of like character are thought of as having enough in common, in spite of great and obvious differences of detail, to be classed under the same heading. In other words, the speech element “house” is the symbol, first and foremost, not of a single perception, nor even of the notion of a particular object, but of a “concept,” in other words, of a convenient capsule of thought that embraces thousands of distinct experiences and that is ready to take in thousands more.
Categorizing an idea lets us feel we understand it by placing it in the company of other ideas that are categorized in the same way. “As soon as the category is at hand, we instinctively feel, with something of a sigh of relief, that the concept is ours for the handling. Not until we own the symbol do we feel that we hold a key to the immediate knowledge or understanding of the concept” (Sapir Reference Sapir1921: 17). A search for this feeling of satisfaction might even sometimes lead us to carry with us a guidebook that will help us experience it, for example with respect to a particular flower or bird.
Categories have various properties that have been described in numerous other works (e.g. Rosch et al. Reference Rosch, Mervis, Gray, Johnson and Boyes-Braem1976; Rosch Reference Rosch, Rosch and Lloyd1978; Lakoff Reference Lakoff1987). They may form a hierarchy such as animal > dog > spaniel > springer spaniel. One of those levels (probably in this case dog) may function as a basic or default level. Certain categories may be favorite or typical representatives of that basic level, as robins are typical birds whereas penguins are not. The fact that an idea may be categorized differently on different occasions shows that categorizing does not force an idea to adhere consistently to a single choice.
An idea may favor a particular categorization to a greater or lesser degree, establishing its degree of “codability” (Brown Reference Brown1958). At one extreme a person may without hesitation characterize a particular idea as an instance of the dog category. At the other extreme a person may have a perfectly clear idea of something while nevertheless lacking a category that will provide it with a sound. Most of us are familiar with the small plastic sheath on the end of a shoelace that allows it to be passed through a small hole. A few people may have learned the word aglet, but probably most of those who are completely familiar with this object have never associated it with that or any other sound. There is a resemblance here to the tip-of-the-tongue experience, but in this case the sound is not just momentarily inhibited but unknown.
Categorizing illustrates repeatedly on a small scale the two components of understanding that were discussed in Chapter 2: observing something particular and inventing something general to accommodate it. We are constantly observing particular events, states, and entities, but our understanding of them would be hopelessly limited if we stopped there. The thousands of categories stored in our minds are thousands of little theories, without which everyday experiences would be a mass of confusion. We can now add categorization as a second step on the path from a thought to a semantic structure (Figure 9.5).
Figure 9.5 Adding categorization to the creation of a semantic structure
Jane interpreted the event she had in mind as an instance of a category whose semantic representation will be shown here as [hang out]. The square brackets are a way of showing that a categorized idea does not, at this semantic stage, consist, for example, of the word hang followed by the word out but is a unitary semantic element directly related to an idea. We have seen also that the idea of this event was associated with the idea of an entity comprising two individuals, Jane and Molly, one of whom was the speaker, and thus this idea was categorized as what can be shown as [first person plural]. The third idea, the location of the hanging out, was categorized in a way representable as [distal], capturing a location some distance from where the speaker was at the moment.
Orientation
We saw in Chapter 5 that ideas are oriented within a multidimensional matrix of time, space, epistemology, emotions, social interactions, and context. It was mentioned that languages differ in the importance they assign to these dimensions, with English devoting considerable attention to time. Figure 9.6 adds orientation to the processes that lead to a semantic structure.
Figure 9.6 Adding orientation
How, then, were the ideas in Jane’s utterance oriented? Mentioned earlier in this chapter was empathy, the ability of a speaker to imagine the thoughts of a listener, and Jane was seen as assuming that two of her ideas – the idea of the event’s participants [first person plural] and the idea of its location [distal] – were already active in Fred’s consciousness because they had been activated in the preceding talk. They were thus oriented as given. In addition, however, the hanging-out event was to be performed, not by Molly alone, but jointly by Molly and Jane, and thus Jane’s [first person plural] was oriented as contrastive (contrasting Molly and Jane as a pair with Molly alone).
Jane assumed that her idea of hanging out was not already active in Fred’s consciousness and that she was about to say something that would activate it; thus it was oriented as new. Other orientations of this event included the fact that it was intended for the future – here we can label it simply future – and that it would require moving to a different location, or andative. How Jane categorized and oriented her ideas is summarized in Table 9.1, showing the idea types, how they were categorized, and how they were oriented.
Table 9.1 Semantic choices in Jane’s utterance
| Idea type | Categorization | Orientation |
|---|---|---|
| event | [hang out] | new |
| future | ||
| andative | ||
| participants | [first person plural] | given |
| contrastive | ||
| location | [distal] | given |
Combination
The ideas of the event, its participants, and its location were not floating independently in semantic space but were combined in a hierarchy of immediate constituents, as explored years ago by Rulon Wells (Reference Wells1947) and subsequently incorporated in some fashion in many theories of language structure. The central idea of hanging out was amplified with [distal], and its participants were added as [first person plural] as shown in Figure 9.7. These three ideas were not temporally ordered, either in thought or in semantic structure, and Figure 9.7 can be imagined as a mobile with parts that swing freely in semantic space.
Figure 9.7 Combining ideas
Figure 9.8 now adds combination as the last step in the creation of a semantic structure from a thought:
From Semantics to Syntax
Semantic choices are seldom passed on directly to phonology but are almost always reformulated in ways that replace semantic structures with what we know as syntax (Figure 9.9).
Figure 9.9 From semantics to syntax
If one views language within the framework developed here, one may well ask why this syntactic stage intrudes itself at this point. Why aren’t semantic structures represented by sounds directly? The answer lies in the fact that languages are always changing, and that innovations in the semantic realm often occur without corresponding changes in already established representations in sound. Those changes fall under the headings of “idiomaticization” and “grammaticalization.” These similar processes create situations in which newly derived semantic elements continue to be represented by the structures from which they were derived. Those earlier structures are now left stranded with “quasi-semantic” elements whose direct relation to thought has been distorted. It is these quasi-semantic elements that proceed to find a representation in sound.
Syntax has something in common with a sailor suit, the traditional uniform of enlisted men in the United States Navy as pictured in Figure 9.10. Parts of this uniform were and still are directly functional in keeping the sailor warm and covering parts of his body that are meant to be covered. But at least two parts of the uniform are no longer directly functional. The jumper flap is there because in earlier times sailors greased their hair and this flap kept the grease from soiling their jumpers. The origin of bell bottom trousers is more obscure. One explanation is that they made it easier for sailors to pull up their trousers from the bottom when they swabbed the deck or climbed the rigging. “Whatever the real reason, reliable documentation validating any assertion was lost long ago. What is true, however, is that the bell bottom trouser has consistently remained as a part of the identifiable occupational dress of sailors throughout modern history” (Dervis Reference Dervis2000). A sailor suit is thus a mixture of functional and no longer functional elements and syntax is like that too, combining elements that are semantically relevant and thus bear a direct relation to thought with elements whose connection to thought has been distorted and whose explanation can only be found in a language’s history.
Figure 9.10 Parts of a sailor suit
How then did it happen that the semantically unitary element [hang out] came to be represented syntactically with the words hang out? The English language is notorious for its phrasal verbs: constructions in which a verb is followed by a particle or preposition. The words hang and out are promiscuous in their availability to such constructions: hang on, hang up, hang around, hang in there, along with bring out, look out, carry out, not to mention nouns like hangout, hangup, and hangover. Jane’s use of hang out seems to have originated in twentieth-century teenage slang as a way of expressing the idea of spending time somewhere without a clearly defined purpose.
The distance between an idiomatic (semantic) meaning and a literal (quasi-semantic) meaning can vary considerably. Jane’s thought lacked the idea of vertical suspension conveyed by hang as well as the container-related meaning of out, but the relation of syntax to thought can be closer, as with let it all hang out. It has been observed that when the relation is relatively close, people may remain peripherally aware of the literal meaning (e.g. Gibbs 1994), or what I have called a “shadow meaning” (Chafe Reference Chafe, Bergs and Diewald2008: 265). The idiom spill the beans has become a canonical example. It expresses the idea of prematurely disclosing a secret, but while that idea occupies center stage one may at the same time experience a visual and perhaps even an auditory image of beans being spilled (Lakoff Reference Lakoff, Caron, Hoffman and Silva1980). The presence and nature of a shadow meaning may vary with the idiom and from person to person, but the point is that quasi-semantic syntactic elements sometimes feed back into the way a thought is experienced without replacing its primary meaning, in this case the idea of disclosing a secret.
We can begin to form a syntactic structure by replacing a semantically categorized idea with the literal form of an idiom, a process called “literalization” in Figure 9.11. The result of this step can be shown by removing the square brackets around [hang out], replacing this unitary semantic element with the words hang and out, and by replacing [first person plural] with the pronoun we.
Figure 9.11 Literalization as a contributor to syntax
Grammaticalization is a process that literalizes the orientation of an idea. It applies here to the orientation of [hang out] as future and andative. Future has by far the more complex history and it provides an excellent example of how far syntax can be separated from semantics as a result of language change. The construction be going to has been discussed frequently as an illustration. Hopper and Traugott (Reference Hopper and Traugott1993) cite the fifteenth-century attestation found in the Oxford English Dictionary: “Thys onhappy sowle … was goyng to be broughte into helle for the synne and onleful lustys of her body.”
This construction, however, could not have arisen if the language had not already acquired the use of be … -ing as a way of expressing the so-called progressive aspect. There is a useful discussion of the history of this construction in A Middle English Grammar by Mustanoja (Reference Mustanoja1960). In Old English there were evidently two constructions. One was a participial construction with the -end ending, the use of which was discussed in Nickel (Reference Nickel1966). The other was a gerundial construction with the -ing ending accompanied by the preposition on. In the thirteenth century the -end ending merged with the -ing ending and the distinction between the two constructions was blurred. Meanwhile the on was reduced to a-, as in he went a-hunting. The form without this a-, simply he went hunting, gradually replaced the other, leaving a-hunting as the quaint relic it is today.
With be … -ing now in place as the way of verbalizing the progressive aspect, statements like I’m góing, to buy bréad (the purpose of my going is to buy bread, expressed in two intonation units) were reinterpreted as a future orientation that was expressed in a single intonation unit without the accent on going. “The change depended on an inference of futurity from purposives” (Hopper and Traugott Reference Hopper and Traugott1993: 3). If I am going somewhere to buy bread, the buying will necessarily take place in the future. English already possessed a future orientation that was realized syntactically with the auxiliary will, as in we will hang out there. The change just described introduced a new and different future orientation with the purposive flavor observable in we’re going to hang out there. The syntactic realization of this now very common future orientation arose in the ways just described, illustrating how syntactic structures can result from a series of semantic innovations.
The syntactic representation of the orientation andative is easier to describe: simply the word go contracted from the sequence go and (Jane might have said we’re gonna go and hang out there). For this andative meaning the English language offers another alternative: go … -ing as in go bowling and go skiing, which appears to be largely restricted to recreational activities that include go shopping. The remaining orientations in Jane’s utterance – given, new, and contrastive – were realized prosodically in ways to be described later. The syntactic structure of Jane’s utterance can now be represented as shown in Figure 9.12, where the andative and future orientations have been literalized as discussed above.
Figure 9.12 The syntactic structure of Jane’s utterance
From Syntax to Phonology
We have now reached the point in verbalization where a syntactic structure on the thought side of language comes to be represented by a phonological structure on the sound side (Figure 9.13). The bottom tier of this figure is the place where thought-based and sound-based representations are linked through symbolization (Chafe Reference Chafe1967; Langacker Reference Langacker1990). The importance that is so often assigned to syntax may stem in part from its location at this point where the thought–sound divide is bridged. The bottom tier is also the place where a direct relation to both thought and sound has been distorted.
Figure 9.13 From syntax to abstract phonology
Here a word or construction relates to both a thought and a sound: to a thought on the left and a sound on the right. For those of us who have grown up in a literate society it may be virtually impossible to think of a word in isolation from its sound or even its spelling. In Chapter 3 Earl could hardly be blamed for associating the word plinth with the way it sounded or was spelled. His conception of that word placed it entirely in the right-hand column of Figure 9.13. Very differently, Harry and Lotte Webster conceived of the Onondaga word onų́:dó:da’ entirely in the left-hand column, with no clear idea of how to spell it and with minimal interest in how it sounded. It was, to be sure, the sound that activated the idea in their minds, but they were no more conscious of that fact than Earl was conscious of the meaning of plinth.
From Abstract to Overt Phonology
Proceeding further along the verbalization path, we can see why it is necessary to distinguish overt from abstract phonology (Figure 9.14). In our example the distinction is realized in the fact that we are is expressed in sound as we’re and going to as gonna. These differences can be attributed again to language change, in this case the contraction of longer sequences into shorter ones. The historical basis of distinguishing semantics from syntax is thus mirrored on the sound side in distinguishing overt from abstract phonology. It is also apparent that sound changes fail to respect the thought side of language. The contraction we’re combines the idea of the participants in the event, we, with are, the first word of are going to, the verbalization of future. The contraction gonna combines only the second and third words of are going to. This independence of sound change from thought-based structures will be demonstrated again in Chapter 10.
Figure 9.14 From abstract to overt phonology
From Overt Phonology to Sound
Figure 9.15 brings us to the end of the verbalization path: the sound that expressed Jane’s thought. Figure 9.16 repeats Figure 9.1 by tracking the fundamental frequency of her sound, where the rising pitch at the end of hang out and the conspicuous fall on there are easily visible. An overall decline in pitch is shown by the dashed line. It can also be seen that Jane’s utterance gradually decelerated, the rapidly spoken syllables of we’re gonna followed by a slower go, a still slower hang out, and finally a much prolonged there. This overall decline in pitch combined with deceleration is characteristic of many, although certainly not all intonation units.
Figure 9.15 From overt phonology to sound
Figure 9.16 The fundamental frequency of Jane’s sound
Figure 9.17 makes the more radical suggestion that prosody and gesture can bypass the bottom tier of Figure 9.15, allowing semantic elements such as given, new, and contrastive to be symbolized directly by prosody, and perhaps these or other semantic elements to be symbolized directly by gesture (Bolinger Reference Bolinger1983; McNeill Reference McNeill2005). In other words, a separate syntactic structure may be irrelevant for certain aspects of prosody and gesture, with gesture resembling prosody in a different form.
Figure 9.17 Semantic features symbolized directly by prosody and gesture
Summary
The brief utterance that was discussed in detail in this chapter illustrated the stages that intervene between a thought and a sound: semantic structure, syntactic structure, and both abstract and overt phonology. We turn now in Chapter 10 to the way a very different type of language would verbalize a very similar thought.
They don’t have articles, and they wouldn’t know how to compensate for this lack of articles either with case or with prepositions, which they also lack. Nevertheless, they have other ways of establishing and maintaining clarity of discourse … They have only a few adverbs and conjunctions, but in fact they have an astonishing wealth of verbs. In their language almost everything is a verb, or can become one.
Cuoq was a French priest who was describing the Mohawk language, which he knew from his missionary work in Canada. Comparing it with his native French and with other languages familiar to him, he found it lacking in the grammatical elements a European linguist would expect a language to have. He was at the same time impressed by the fact that its speakers had no problem, in spite of what might have seemed major handicaps, in “establishing and maintaining clarity of discourse,” noting that they compensated for what they lacked with “an astonishing wealth of verbs.” No better illustration could be found of the different paths languages may follow as they move from a thought to a sound. In this chapter we trace that path in the Seneca language, a close relative of Mohawk.
Exact translations are never possible, and the thought expressed by Jane in Chapter 9 as we’re gonna go hang out there cannot be reproduced exactly in Seneca. Discussed in this chapter, however, is a reasonable approximation. Seneca of course lacks the idiom hang out, which is replaced here with a Seneca idiom that can be translated visit. Unlike Jane’s utterance this example was not taken from spontaneous speech but it is a realistic thing to say, even though in actual speech it would almost certainly be accompanied by several particles: short words that would position it within the ongoing discourse. Let us suppose that it was uttered by a person named Andrea, who said:
Hëyagyajö́’se:nö’.Footnote 5
“We will go visit there.”
Immediately striking is the contrast between this one Seneca word and the five words of the English. It has often been said that a polysynthetic language expresses in one word what a language like English would express in an entire sentence. It thus becomes necessary in this chapter to trace the path from a semantic structure to a morphological structure, the internal structure of a word. The term “morphosyntactic” in Figure 10.1 leaves room for both morphology and syntax, and later in this chapter we will see how both may be relevant.
Figure 10.1 From a thought to a sound by way of morphosyntax
Andrea’s Thought
We saw in the previous chapter that Jane’s thought included the idea of two participants in the hanging-out event, Jane and Molly, verbalized with the word we. Chapter 5 mentioned how ideas of particular events necessarily include ideas of entities that initiate them, are affected by them, or otherwise participate in them. That observation applies to all languages, but it is realized directly in a polysynthetic language, where the participants in an event are tightly packaged together with the event itself within a single word.
If a Seneca verb corresponds to an English sentence, one might at first suppose that the range of thoughts expressed by such a verb would rival the range of thoughts that can be expressed in an English sentence. But that is not the case. A Seneca verb does make it possible to verbalize a number of things that are relevant to an event, including not only its participants but also its aspectual properties, its location in space and time, whether it is distributed across several locations, its relation to reality, whether it is a cause or an instrument, whether it benefits someone, whether it involves going somewhere to do something, whether it is performed while walking, whether it happens easily, or whether it is the culmination of a series of other events. Selections from these numerous possibilities are packaged tightly within the verb, where each element has a form and position determined by the language’s morphological patterning.
But this list of morphological possibilities is only a small subset of the open-ended opportunities that are offered by combining words in sentences, and later in this chapter we will see how the nearly infinite variety of possible thoughts can lead speakers of a polysynthetic language to supplement their morphology with options provided by their syntax. But it is still worth asking whether the thoughts of a Seneca speaker have at least a tendency to conform to the options offered by the language’s morphology as listed in the previous paragraph. Does the easy availability of those options bias a speaker toward thinking in those terms? And do those options feed back into the ways Seneca speakers think? The larger question of whether speakers of different languages think differently will be confronted in Chapter 13, but for now we can focus on the present example.
Most Seneca words are built on a base that offers a large number of lexical possibilities. This base is usually accompanied by one or more participants and also by a mode and aspect. In our example Andrea’s thought is likely to have contained at least the three ideas on the left in Figure 10.2, where placing them in a single column shows their simultaneous presence within the word, as compared with Jane’s thought on the right as copied from Figure 9.2, where they were distributed across three words.
Figure 10.2 Andrea’s thought compared with Jane’s
Andrea’s Semantic Structure
Andrea’s semantic structure had as a base her categorization of the event as [atyö’se]. This notation follows the convention introduced in Chapter 9 where Jane’s semantic category was shown as [hang out], the square brackets showing here that [atyö’se] is a unitary semantic category and not at this stage composed of the separate parts described below. The participants in this event, corresponding to Jane’s [first person plural], can be shown as [exclusive dual agent]. But whereas [first person plural] was assigned in English to the separate word we, the Seneca participants were included in the same word as the event and will thus be represented directly below it, as in Figure 10.3. “Exclusive dual” means that the participants excluded the listener (Fred) and that there were two of them, Andrea and Molly. “Agent” means that those two would be the instigators of the event. Seneca also requires that an event be specified in terms of mode and aspect, shown here as [predictable event].
Figure 10.3 Elements of Andrea’s semantic structure
The idea of this event was semantically oriented as shown in Figure 10.3 by placing andative and distal without brackets below the three ideas. The andative orientation functions as it does in English: the event required moving to a different location. The event was also oriented as distal, or located at a distance from the speaker. This orientation contrasted with the separate idea that was expressed in English with the word there.
The properties just described – a base, its participants, and its mode and aspect – are obligatory for most Seneca verbs, and in this case they were supplemented with the andative and distal orientations.
An important question arises at this point. English does not say explicitly that there were two people and that the listener was excluded, but to what extent are people conscious of knowledge that is not directly verbalized in their language? In this case Jane obviously knew that the participants in the hanging-out event were two people and that neither of them was Fred, even though that knowledge was not represented in the sound she uttered, whereas it was represented in Andrea’s sound hëyagyajö́’se:nö’. We saw in Chapter 9 that a thought nearly always includes more than is expressed in language, but the question now is whether Jane’s knowledge that there were two people going to Steve’s, neither of whom was Fred, occupied a less active state in her mind than the corresponding knowledge in Andrea’s mind, where it was expressed overtly in the sound she uttered.
In my own imperfect attempts to speak Seneca I have sometimes found myself using a plural form of a Seneca verb when I should have used a dual, even though I knew that only two people were involved. Introspection suggests that it would have required an extra cognitive effort for me to pay attention to the fact that there were only two people. It was something I knew, but that knowledge was not as active in my consciousness as it would have been for a fluent Seneca speaker. Observations like these suggest the value of distinguishing Jane’s tacit knowledge from Andrea’s overt knowledge (cf. Polanyi Reference Polanyi2009). How the brain distinguishes overt, verbalized knowledge from tacit knowledge is a question that clearly deserves further study.
Andrea’s Morphological Structure
How were these semantic elements represented morphologically? Just as the semantic orientation future acquired its syntactic representation as be going to in the historical stages described in Chapter 9, the morphological representation of [atyö’se] can also be traced to its history. First it came to be represented with a verb root meaning “arrive” accompanied by a “benefactive” derivation. In other words “visit” was expressed literally as “arrive for someone’s benefit.” That idea, however, was transitive (visit somebody) and in our example the visiting was intransitive, a fact that was expressed by adding a marker of “middle voice” (e.g. Kemmer Reference Kemmer1993). The meaning “exclusive dual agent” was represented by an element meaning “exclusive agent” to which “dual” was added separately. These changes led to a morphological structure that contained the elements shown in Figure 10.4.
Figure 10.4 Elements of Andrea’s morphology
What was then passed on to phonology is shown in the linear template above the horizontal line in Table 10.1. The items above the line belong to the thought-based face of morphology, those below the line to the sound-based face. Removing the spaces from the elements below the line gives us the reconstructed phonological form hëyakyatyö’sehna’.
Table 10.1 A verb template with morphology above and phonology below the line
| Distal | Future | Exclusive agent | Dual | Middle | Arrive | Benefactive | Andative | Eventive |
|---|---|---|---|---|---|---|---|---|
| h | ë | yak | y | at | yö | ’se | hn | a’ |
From Reconstructed to Surface Phonology
Through both internal and comparative reconstruction but also through the fortuitous existence of French–Seneca and Seneca–French dictionaries from about 1700, we know quite a bit about the state of this language 300 years ago. During the centuries since then the language underwent more than fifty sound changes, most of them during the eighteenth century. As a result the difference between a reconstructed phonological structure and the sound of a word today can be considerable. The following changes include only those that are relevant to this example:
| Reconstructed form: | hëyakyatyö’sehna’ |
| Accent the fourth syllable: | hëyakyatyö́’sehna’ |
| Delete h before n and lengthen the preceding vowel: | hëyakyatyö́’se:na’ |
| Change k to g and t to d, before y: | hëyagyadyö́’se:na’ |
| Change dy to j: | hëyagyajö́’se:na’ |
| Change a to ö after n: | hëyagyajö́’se:nö’ |
The last line contains the form of this word that is now realized in sound, and it is the way Andrea expressed her thought.
The stages described above show how a linguist might account for the steps that led from Andrea’s thought to the sound she uttered, but there is no reason to suppose that Andrea herself followed those steps. To a large extent Seneca speakers remember words as wholes, not as the morpheme sequences identified by linguists. They are aware of the meanings that are contained within a word, but they are often unable to associate parts of a word with those meanings. If that is the case, how do they manage to create a word they have never before spoken or heard? In fact they do that less often than one might suppose. Verbal creativity is an ability that varies with different speakers, and creative eloquence is admired. A creative speaker almost certainly does not follow the stages described above but relies instead on what might be termed a “folk morphology” consisting of ad hoc patterns established individually and analogically on the basis of words that are already familiar.
Establishing and Maintaining Clarity of Discourse
At the head of this chapter Father Cuoq was quoted as saying that a language of this type was able to establish “clarity of discourse” by taking advantage of its “astonishing wealth of verbs.” We can end this chapter by looking at one of the ways that is accomplished. Despite the wealth of resources offered by Seneca morphology, speakers often find them inadequate to express what they are thinking. In that case Seneca syntax makes use of a pattern that allows the information within a verb to be amplified in order to convey more of what is included in a thought. As an example we can start with a thought that might have been expressed in English in this way:
There were a father and son who came to visit.
We can compare that English sentence with the following Seneca:
| Né:ne:’ | wa:ya:jö’s | neh, | yadátawak. |
| those | they (masculine dual) visited | namely | a father and his son |
| “They visited: a father and his son.” | |||
The verb wa:ya:jö’s captured the idea of visiting by two males. But that verb failed to convey everything the speaker was thinking, so he went on to add the idea that the two visitors were a father and his son. The third Seneca word, the particle neh, signaled that the information in the verb was about to be amplified by what would follow. The translation “namely” captures this function reasonably well, although in writing one might simply use a colon, as followed the word visited in the free translation here. Most Seneca kinship terms are verbs, and the final word, yadátawak, can be translated literally “they were father and son to each other.” This amplification repeats a reference to the participants who were included in wa:ya:jö’s. The y- in both verbs showed that the participants in both the visiting event and the father–son relationship were two males, thus explicitly linking the two ideas. Amplification constructions like this are a common feature of Seneca syntax (Chafe Reference Chafe2012, Reference Chafe2015), a common way of establishing “clarity of discourse.”
One sometimes reads that children learn approximately 14,000 words by the age of six. Many of those statements are traceable to studies that were compiled in a single source (Templin Reference Templin1957). Those studies mentioned the problem of deciding whether, for example, look, looks, looked, and looking should count as one word or four. Presumably children learn the meaning and form of the lexical item look, but also learn the inflections that can be applied to a large number of such items. Irregular items like go and went can of course only be learned separately. With Seneca the number of possible inflections of a single verb base is greatly magnified, reaching into the hundreds, and irregularities are extensive, with cases resembling go and went on a larger scale. For a child learning Seneca, much depends on learning a very large number of individual words at an early age and deriving from them ad hoc patterns of a folk morphology, each such pattern being applicable to a restricted number of cases.
Summary
Seneca served here as an example of the polysynthetic language type, packaging within a single word a variety of elements resembling those which in English would be assigned to separate words. Beginning with a thought that combines what in English would be separate ideas, it categorizes and orients that thought in its own ways, which leads to a morphological structure in which those elements are arranged in a linear verb template. The result is then subject to a number of phonological changes that reflect the language’s history during the past 300 years. Although a variety of options are available for inclusion within the meaning of a verb, they are often not enough to cover everything a speaker is thinking, in which case recourse can be had to a syntactic construction that amplifies the meanings within a verb in order to express more of what is contained in a thought.
