2. Myth and music in linguistic relativity

Over the years, Friedrich (Reference Friedrich1979, Reference Friedrich1986, Reference Friedrich, Jourdan and Tuite2006) has attempted to reformulate linguistic relativity with an emphasis on poetic language (see also Hymes Reference Hymes1981, Sherzer Reference Sherzer1987, Becker Reference Becker1995; for a review see Leavitt Reference Leavitt2011). As Friedrich (Reference Friedrich1986: 43) has noted, “poetic language is the locus of the most interesting differences between languages and should be studied together with the poetic imagination of the individual.” Friedrich takes inspiration from this famous quote from Edward Sapir (Reference Sapir1929a: 209):

Elsewhere Sapir (Reference Sapir1921: 225) notes that “every language is a collective art of expression. There is concealed in it a particular set of esthetic factors – phonetic, rhythmic, symbolic, morphological – which it does not share with any other language.” This means that the dialogue between languages can create particularities of aesthetic features of sound and rhythm, symbol and morphology.

Such concerns with poetry can also be found in the work of Franz Boas (Reference Boas, Boas and Powell1966: 58):

For Friedrich (Reference Friedrich1986: 35), the question of linguistic relativity hinges on the relationship between myth and the music of language or, as he describes it:

The music of language is then intertwined with what Friedrich (Reference Friedrich1986: 37) calls myth. For Friedrich, myths are our “ideas and theories” that underlie our imagination, as well as what are conventionally considered myths in anthropology, and our own “myths of the unique individual” (Friedrich Reference Friedrich1986: 38). Friedrich's definition of myth is meant as a heuristic that captures the complex interplay of ideas and feelings about and of the world.

In this article, I will be most concerned with the music (or sound) of this poem and how that influences the ways that it is heard. This work builds on the theoretical scaffolding developed in Webster (Reference Webster2014b). As John Leavitt (Reference Leavitt2011: 202) has noted, “phonetics and phonology represent unavoidable sound-worlds.” It is here, I would argue, that “the specificity of language is most intensely” felt (Leavitt Reference Leavitt2011: 202). A number of years ago Dwight Bolinger (Reference Bolinger1940: 61) observed “that certain words naturally go together, and being encountered drag each other along a train of thought like the barbed atoms of Democritus.” They are natural only insofar as we perceive such resemblances as socialized language users. Bolinger went on to list such sound associative practices as alliteration, rhyme, homonyms and puns. He called such sound associations “word affinities.” In arguing that words were not arbitrary – a point to be taken up again by Friedrich (Reference Friedrich1979) – he noted that languages are systematic. Such a view, of course, recognizes Sapirian leakage and the fact that while systematic, languages are not discretely bounded billiard balls, they are not self-contained a priori systems. Languages hang together and come together through and in use (see Hopper Reference Hopper, Aske, Beery, Michaelis and Filip1987, Reference Hopper1996).

In a later article, Bolinger (Reference Bolinger1949: 55) argues that, “when we speak of sound-suggestiveness, then, we speak of the entire language, not just of a few imitative or self-sufficient forms.” Stated more poetically, Bolinger (Reference Bolinger1949: 56) suggests that “the phonetic elements of a language are like the keys of a piano. They have been played so often and in so many combinations that even a random chord, struck by an object accidentally falling on them, will have some vague semblance of meaning.” This, then, is the fire that lurks in sapphire for Bolinger (Reference Bolinger1940). This is the relatively non-arbitrary nature of the symbol that Friedrich (Reference Friedrich1979) describes. As Dell Hymes (Reference Hymes and Sebeok1960: 112) makes clear, “if we are to understand a fair part of linguistic change, comprehend the use of language in speech and verbal art, take account of all the varied speech play in which a competent speaker may indulge, and to which he can respond, we must study his real and lively sense of appropriate connections between sound and meaning.” Hymes here should remind us that we cannot ignore myth because it cannot be completely separated from the music of language.

The questions that inform this article, then, revolve around the following: Does it matter what the sounds of a language are? What the sounds of various words or morphemes are in a language? How such sounds resonate within a language? How such sounds convoke (that is, call together words and the sounds of words)? As Leavitt (Reference Leavitt2011: 210) describes in his review of the history of attending to linguistic relativities, the Boasians – Boas, Sapir, and Whorf here – did consider sound to be of importance (see also Urban Reference Urban1986). It mattered to Sapir (Reference Sapir1929b) and also to Whorf (Reference Whorf and Carroll1956: 266–268), concerned, as they were, with expressive devices and phonetic symbolism. Boas (Reference Boas1889) famously noted that the sounds of a language might predispose one to hear the sounds of an unfamiliar language not on the terms of the unfamiliar language, but rather in terms of the language most familiar. Interlingual puns, I might add, play on just such a tendency. To hear, that is, god as gad ‘juniper tree.’ But to bring up punning in a discussion of linguistic relativities might seem too much. Such a view is conditioned, as our modernist conceits so well described by Bauman and Briggs (Reference Bauman and Briggs2003) and Samuels (Reference Samuels2004) suggest, on the low standing of puns in the serious work of language. But punning is, after all, and to borrow the term from Samuels (Reference Samuels2001), a form of phonological iconicity. Words resemble other words through sound (they are bivalent, in Woolard's Reference Woolard1998 terminology). We will see this concern with punning to be an important aesthetic in the poetry of Jim. Much of the current debates about linguistic relativity – informed by modernist visions of what a language is or should be – take a narrow semantico-referentialist view of language (Bauman and Briggs Reference Bauman and Briggs2003, Samuels Reference Samuels2004), imagining that (spoken) languages lack phonetic substance (Webster Reference Webster2015c). It seems to me that we should attend to sound suggestiveness via phonological iconicity and the ways sounds evoke and orient our imaginations as a crucial locus for thinking through linguistic relativities. To be clear, by linguistic relativity I mean a positive vision of the ways in which languages facilitate possibilities to orient and imagine (see Friedrich Reference Friedrich, Strecker and Tyler2009: 219).

3. The poem and its translations

The poem that forms the core of this article comes from Rex Lee Jim's (Reference Jim1995) all-Navajo collection saad. The volume was published by Jim's alma mater, Princeton University, through its Princeton Collections of Western Americana. I have discussed this book and translations of other poems from it elsewhere (Webster Reference Webster2006, Reference Webster2013, Reference Webster2015a; Mitchell and Webster Reference Mitchell and Webster2011); here I note that the book is written almost entirely in Navajo (including page numbers and introductory materials). I should also note that most Navajos, given the general lack of literacy in Navajo, come to this poem as an oral text and not as a written text (Mitchell and Webster Reference Mitchell and Webster2011). The consultants that I have worked with on this poem for this article are all literate in Navajo. Here is the poem by Jim (Reference Jim1995: dízdiin dóó bi’ąą tseebíí):

• éh

• tsidił ga’

• da'diłdił

• yiits'a’

Here are the four translations.

Translation 1 was done by Blackhorse Mitchell (Navajo poet, writer, singer, performer, sheepherder, and educator). Translation 2 was done by a Navajo consultant who wishes to remain anonymous, but has worked as a translator for other anthropologists over the years. Translation 3 is by a mutual friend of Mitchell and me. He has not done much translation work. Translations 1 and 2 were done face-to-face on and around the Navajo Nation in 2008. Translation 3 was done via text messaging in early 2014. Translation 4 was done by me in the spring of 2014 as well. Let me be clear here: I do not claim that any of these translations is right or wrong. I think they all say something interesting about this poem. To say that one translation or another is a mistake assumes a kind of singularity of meaning for Jim's poems. However, in working on a number of poems by Jim, I've come to realize that ambiguity, bivalency and indeterminacy are important components of his poetry (I will return to this point at the end of this article). Nor do I want to leave the impression that I consider my translation to be the final say on the matter. Each translation facilitates a particular glimpse of this poem. It should be clear that all these translations engage in exuberances and deficiencies (Becker Reference Becker1995). They leave out too much, put in too much. One thing they leave out is the sounds of the poem in Navajo. New sounds are introduced in the English translations.

4. The Morphology of a Poem

I begin with a traditional morphological analysis of the poem, to be followed by a more detailed discussion of the salient features of the morphology in a bit more detail. The poem can be morphologically analyzed as in (1).

1. (1) éh

   intrj

   tsi(n)- -dił ga’

   wood shake by concussion emph

   da' di-  -ł- -dił

   dist thematic prefix for elongated object caus slender object moves

   yii- -ts'a’

   thematic prefix for sounds existing to sound

The first line consists of the interjection or exclamation éh, glossed into English by my consultants as “wow,” “oh,” “what do you know,” and “yes” (this last form was given in conversation about the translation). From this perspective, there are no ideophones in this poem.

Many of Jim's poems create an intensification of form through the repetition of a key sound in that poem (in this poem, that sound is -dił-) (Webster Reference Webster2013, Reference Webster2015a; on intensification of form, see Jakobson Reference Jakobson and Sebeok1960). The second line introduces that key sound: tsidił ga’ can be analyzed as tsin ‘wood’ + -dił ‘jar, shake by concussion’ (-déél ‘jar, shake by concussion, play stick dice’ Young and Morgan Reference Young and Morgan1992: 134, 135). This form is often translated as “stick dice” (Aberle Reference Aberle1942: 144, Young and Morgan Reference Young and Morgan1987: 25, Wall and Morgan Reference Wall and Morgan1994: 152). The Franciscan Fathers (Reference Fathers1910: 482) translate the form as ‘stick rebounds’. In my own translation, I have taken my cue from the Franciscan Fathers here in translating line 3. ga’ is an emphatic particle that particularizes nouns (Young and Morgan Reference Young and Morgan1987: 22). Given its low salience on the cline of metalinguistic awareness, it often was not translated by consultants (see Silverstein Reference Silverstein and Duranti2001).

As for the third line da'diłdił; da’ seems like the distributive prefix here or perhaps also the interrogative da’ ‘is it’; di- + -łdił ‘to cause them to move independently’ (see, for example, ‘adishdił ‘to play stick dice’ (Young and Morgan Reference Young and Morgan1987: 25)). Young and Morgan (Reference Young and Morgan1987: 25) go on to explain that “di-, Pos. VIa, generic classifier relating to an elongated object? + *łdił: to cause it or them to move independently – a slender flexible object, as a rope or pair of objects, as shoes, stick dice.” Blackhorse Mitchell, in a conversation about this form, said, “da'diłdił is like throwing the stick repeatedly.” The sense of repetition suggested by Mitchell is possibly a result of the distributive da’ prefix. Reichard (Reference Reichard1951: 135) notes that this form “may denote a repetitive idea.”

The final line, yiits'a’, can be analyzed as yii- ‘3sg’ + -ts'a’ ‘having to do with sound’ (Young and Morgan Reference Young and Morgan1992: 625). This form, as I have noted elsewhere (Webster Reference Webster2009), is often used as a verb of sounding following an ideophone.

With this background, let us look at the translations again. All four translations recognize the exclamatory or interjectional quality of éh. Mitchell translates the form as wow, while Translator 2 and I go with oh. Translator 3 essentially lexicalizes the exclamation as “what do you know.” All indicate a sense of surprise here. The translations begin to diverge in line 2. Mitchell, Translator 2 and I all remain faithful to the four-line structure of the poem and attempt to translate line by line. Translator 3, on the other hand, gives the gist of the poem in line 2. Mitchell, Translator 3 and I recognize the form tsidił as “stick dice” or “stick game.” I will discuss the stick dice in more detail in Section 6; for now, it is important to note that the stick dice game is a traditional Navajo practice and that it is played by way of throwing long sticks against a rock. Translator 3, seeing tsidił as an important component of Navajo culture, decides against translating the form into English (this is a not uncommon strategy – see Webster Reference Webster2009). Mitchell attempts to capture the particularizing function of ga’ with the, while I use these. On the other hand, Translator 2 does not translate this form as “stick dice” or “stick game,” but rather as the sound symbolic bang. What's going on here?

Before we can answer that question, let me say something about the process of translation for each of our translators. For Mitchell, I mailed him a typed version of saad and he then went about translating the poem from Navajo into English based on his knowledge of the poem. He read the poem aloud numerous times and thought long about the poem before engaging in a translation. Then, in the summer of 2008, I stayed with Mitchell at his home while we worked on a number of translation projects. I audio-recorded our work together (see Interlude). He is literate in Navajo and has taught Navajo at both the high school and college level. He also has written his own poetry in Navajo. Mitchell clearly saw his task as translating poetry into poetry. Translator 3 is also literate in Navajo and he translated the poem after I texted it to him in the middle of a conversation we were having about poetry. This was the first time I had had him translate a poem for me, and he did not translate it line by line; rather he translated the gist of the poem. He was also reluctant to translate tsidił into English – both because it is, for him, an important part of Navajo culture and because he assumed (rightly) that I knew what the form meant. Translator 2, on the other hand, had not worked with me previously, though we had met before. I had heard about his facility with translation from another anthropologist and so sought him out. We met at a local coffee shop in a border town. I had brought two typed copies of Jim's saad with me (one for him to write on and one for me to make notes on). I also audio-recorded our work that day (it took us a little over two and a half hours to translate saad). His process in translating the poems was to read the poem aloud several times and then give an English translation. I would then ask him various questions about the translation. He often wanted to explain the poetic work being done in Navajo to me, but seemed less concerned about translating poetry into poetry.

Translator 2 read line 2 both as ts'idil and as tsidił. He settled on reading it as the first form. That form is analyzable as ts'i- a thematic prefix having to do with sounds (Young and Morgan Reference Young and Morgan1987: 359; Young and Morgan Reference Young and Morgan1992: xiii) and the ideophone dil ‘rumble, stomp, bang.’ The differences between the two readings of line 2 are an ejective alveolar affricate /ts’/ versus a plain alveolar affricate /ts/ at the beinning of the word, and a voiceless lateral (“barred l”) /ł/ versus a voiced lateral /l/ at the end. The corresponding semantic difference is between the stick game and a banging sound.

In line 3, Mitchell – with stomping – and I – with rebounding – attempt to evoke something of the sound of the game being played. I also attempt to capture something of the repeated nature of the sound as suggested by Mitchell in conversation and in the use of the da’ prefix here by reduplicating rebounding. Translator 3 leaves this largely untranslated. Translator 2, however, translates this line as bang, bang. In listening to the audio recording, it seems clear that he has settled on hearing this form as da’ dil dil. In conversation he translates da’ as the interrogative “you mean?” but leaves that form out of his final translation. He translates dil dil as ‘bang, bang’ – again treating this as the sound symbolic form dil and here reduplicated (a common practice in Navajo ideophony – see Webster Reference Webster2009). Translator 2 described this sound as “like an automobile with a rackety back with stuff banging around.”

In line 4, all four translators attempt to get at the soundedness of what has occurred. Mitchell translates it from the first-person perspective as I hear, while the other three translate the form as it sounds.

5. Interlude: A Conversation with Blackhorse Mitchell

Before moving to a discussion of the seduction of ideophony, I will first present an example of how the work of translation is done. Below is the transcript of a conversation that Mitchell and I had about the translation of this poem on July 18, 2008. We were seated in the dining room of Mitchell's home outside of Shiprock, NM on the Navajo Nation. Mitchell had translated the poems for me prior to my arrival on the reservation earlier that month and we were now working through his translations. I would ask for clarifications and the like about his translation. We would then talk through his translations. It was not the only time we talked about this poem, but this example was our first run through the poem together and suggests how the work of translation proceeds. I should also note that I had not translated this poem with Translator 2 or 3 yet. Lines have been organized according to breath pause, and attempt to reveal the cadence of our conversation, with a colon (:) used to indicate phonetic lengthening. Brackets indicate contextual information. Lines have been numbered for ease of reference.

7-18-2008. Mr. Mitchell's home [begins 20:50]

• BM: this one says éh 1

• tsidił it's the stick game 2

• you know those stick thing you use  3

• that's called tsidił 4

• AW:  mhm 5

• BM:  it says tsidił ga' 6

• it says the stick game 7

• was or is 8

• AW: ga'  9

• BM: ga'  10

• AW: was  11

• BM:  ga’ something like that 12

• it says da'diłdił 13

• da'dił 14

• dił  15

• says da'diłdił 16

• AW: da'diłdił 17

• BM: da'diłdił  18

• diłdił is like you're, you're, you're, you're repeating like you're just 19

• it's like the stick game is itself 20

• going like  21

• [hand gestures of throwing or bouncing of stick dice]

• AW: mm  22

• BM:  that's what diłdił is 23

• AW:  and that's the sound it's making 24

• BM:  no  25

• AW:  no  26

• BM: I would, I would 27

• yeah da'dił diłdił yiits'a’ 28

• if you say dil dil yiits'a’ would be like somebody 29

• [stomps feet repeatedly]

• AW: mhm 30

• BM:  that's making dil dil dil dil yiits'a’ 31

• not diłdił 32

• he's making it similar too here 33

• da'diłdił would be: 34

• the stick game itself is playing the game 35

• AW:  but that doesn't evoke the sound 36

• BM: that would be dil yiits'a’ 37

• AW: what does da'diłdił mean 38

• BM: da'diłdił is like throwing the stick repeatedly 39

• AW:  throwing the stick repeatedly 40

• BM: yiits'a’ is its sound  41

• [5 sec pause]

• AW: but then yiits'a’ doesn't change that into the sound it's making 42

• BM: it could be both 43

• it could be dil dil 44

• or da'diłdił 45

• AW: okay 46

• BM: see diłdił 47

• dił by itself is blood 48

• you could say blood, blood 49

• AW:  right 50

• BM: but in this case it's not blood, blood 51

• AW: right 52

• BM: it's a: 53

• da'diłdił would be just like leaping or prancing higher than the leap

• that would be diłdił 54

• so it's got three meanings dil dil 55

• it could be the sound 56

• it could be the leaping motion whatever that is 57

• yiits'a’ makes it its sound 58

• AW: mhm 59

• [ends 23:30]

I will have more to say below about this fragment of conversation, but I have reproduced it here to give a sense of the progression of our conversation about this poem.

6. Seduced by Ideophony

To begin to think about Translator 2's translation of this poem, I turn briefly to recent work on ideophony. Specifically, Mark Dingemanse (Reference Dingemanse, Michelucci, Fischer and Ljungberg2011, Reference Dingemanse2012) has suggested that there can be a certain coercive power to iconicity (following Charles Sanders Peirce, a resemblance of something to someone – see Silverstein Reference Silverstein, Basso and Selby1976, Samuels Reference Samuels2004, Webster Reference Webster2009). Dingemanse (Reference Dingemanse2012), for example, points to the ways that some can see all Chinese characters as pictorial or all ideophones as equally imitative (iconic), even though not all Chinese characters are pictorial, nor are all ideophones equally imitative (iconic). While we tend to assume that ideophones are inherently iconic, Dingemanse (Reference Dingemanse, Michelucci, Fischer and Ljungberg2011: 51) suggests that “the depictive nature of the ideophone coerces us into treating the word as an adequate rendition of the depicted event.” In a later piece, Dingemanse (Reference Dingemanse2012) goes on to argue that:

Almost all Navajo ideophones are, in fact, onomatopoetic (Webster Reference Webster2009). So the question here is not whether or not a peripheral set of ideophones are being understood as sound-imitative or iconic. Instead, what I find most useful in this discussion is the question of framing and how that might “coerce” one into hearing an ideophone. It should also be clear, given what I've described about Navajo ideophony (Webster Reference Webster2009, Reference Webster2014a) and what Janis Nuckolls (Reference Nuckolls1996, Reference Nuckolls2010, Reference Nuckolls2014) has described for ideophony in Runa, that there is nothing particularly “simple” about sound-imitative ideophones when looked at from the perspective of poetics and their place within cultural frameworks of meaning and moral responsibility, though it should be clear those frameworks of meaning and moral responsibility as expressed by and through sound-imitative ideophony are not identical among Navajos and Runa.

It should also be clear that we are dealing with a felt iconicity (Webster Reference Webster2009, Reference Webster2014a; see also Sicoli Reference Sicoli2014), the ways that forms are felt to resemble something, and such felt iconicities are often language particular or, in the case of interlingual puns, interlanguage particular (Webster Reference Webster2010, Reference Webster2013). All forms in a language, it seems to me, have the potential to be felt as iconic of something (see Bolinger Reference Bolinger1940, Reference Bolinger1949). Take, for example, what is not a form that is conventionally understood as ideophonic or iconic in Navajo (tó ‘water’), but can be used – especially by a good poet – as an ideophone for the sound of water dropping. Sherwin Bitsui, for example, in a poetry performance in Edwardsville, IL on November 17, 2009, begins a selection from his book Flood Song (Reference Bitsui2009) in the following manner (lines are organized by breath pause):

1. (2) tó

   tó

   tó

   tó

   tó

   tó

The repetition of tó ‘water’ in such a rhythmic manner evoked the sound of water dropping on the ground. This was, indeed, as he explained to me later, part of the aesthetic that he was after in his performance. In use, then, the repetition of tó created a felt iconicity to the dripping of water.

As I have described elsewhere, Navajo ideophonic expressions (what Navajos sometimes call hodiits'a’ ‘there is a sound’) are often followed by a verb of sounding based on the verb stem -ts’ᶐˊᶐˊ’ (and its suppletive neuter stem form -ts'a’) (see Young and Morgan Reference Young and Morgan1992: 625, Webster Reference Webster2009). Below is an example from a narrative told by Curly Tó Aheedlíinii to Father Berard Haile concerning Porcupine and Elk (1984: 104). Porcupine is trying to discern whether or not Elk – who has ferried Porcupine across a river inside its stomach – has reached dry land. I give this example to show how the verb of sounding works and to give an example of the ideophone dil dil (bolded below) in use (the Navajo is from Haile Reference Haile1984; the English translations were done in consultation with Mitchell).

1. (3) Nikináázdeestał, ńt’éé’ dil, dil, yiists’ ᶐˊᶐˊ’ jiní

   She stamped her feet again, then, dil, dil, it sounded, they say.

   T’áadoo tó diists’ ᶐˊᶐˊ’ da jiní

   There was no sound of water, they say.  Haile (Reference Haile1984: 104)

The context of translating a number of poems by Jim also matters. In an earlier poem in saad, Jim does use the verb of sounding (yiits'a’ + the subordinating enclitic –go) after a reduplicated ideophone (ts’ǫǫs, ts’ǫǫs). I have discussed this poem at length elsewhere (Webster Reference Webster2006, Reference Webster2014a), so here merely give the Navajo and an English version. The ideophones are shown in boldface.

1. (4) na'asts’ǫǫs             mouse

   ts’ǫǫs, ts’ǫǫs              suck, suck

   yiits'a'go                  sounding

   ííts’ǫˊǫˊz                  kiss

   (Jim Reference Jim1995: 37)              (Webster Reference Webster2006: 39)

My point is that Translator 2 was, in a sense, open to hear ideophones in Jim's poetry because he had previously encountered such forms.

More than that, however, I think the verb of sounding yiits'a’ seduced Translator 2 into hearing the poem as a poem composed of ideophones. Here I differ with Dingemanse (Reference Dingemanse2012) and do not see this framing so much as a form of coercion (which seems predicated on a kind of force; you must do this), but rather as seduction (which seems more about enticing and charming; you want to do this). When Translator 2 first read the poem aloud he pronounced the key forms as follows: tsidił ga’ and da'diłdił. However, after he read the final line, he went back and pronounced those forms as ts'idil and da’ dil dil. I think the reason that Translator 2 went back and reread these forms the way he did was because he had been seduced by yiits'a’ into hearing this poem as full of ideophones. The verb of sounding yiits'a’ reframed this poem as an ideophone poem for Translator 2. In this respect, it became an ideophone poem similar to the ones described by Philip Noss for Gbaya (Reference Noss, Erhard Voeltz and Kilian-Hatz2001: 266–267; see also Barrett Reference Barrett2014 on Mayan ideophone poems). There are, I might add, also song traditions that are overwhelmingly composed of either vocables or ideophones among Navajos (see Matthews Reference Matthews1889, Frisbie Reference Frisbie1980).

Mitchell too, I might add, in the conversation reproduced above, recognizes the seductive power of the verb of sounding, to rehear the poem as full of ideophones. After first disagreeing with my question about whether or not the form evokes the sound of the stick dice game being played (lines 24–26), Mitchell then begins to work through the implications of the form in the following excerpt:

• BM: I would, I would 27

• yeah da'dił diłdił yiits'a‘ 28

• if you say dil dil yiits'a’ would be like somebody 29

• [stomps feet repeatedly]

• AW: mhm 30

• BM:  that's making dil dil dil dil yiits'a‘ 31

• not diłdił 32

• he's making it similar too here 33

• da'diłdił would be: 34

• the stick game itself is playing the game 35

Having worked through a number of Jim's poems and recognized a poetic device, Mitchell, in line 33, notes that “he's making it similar too here.” The poetic device that Mitchell has recognized is what is sometimes called saad aheełt’éégo diits'a’ ‘words that resemble each other through sound’, or what we might call punning (see Webster Reference Webster2013). The poetic work of the use of punning in Jim's poetry was best encapsulated by a Navajo who had been listening to Jim read his poetry in December 2000 and remarked, “when you think about it [the sounds in the poem], the whole meaning changes.” Indeed, it was Mitchell's insight that the consonant cluster -chx- used in one poem in saad evoked a concern with hochxǫ’ ‘disorder, ugliness, lack of control’ that led us to co-author an article together (Mitchell and Webster Reference Mitchell and Webster2011). Nor is this practice unique to Jim or to poets; Barre Toelken and Tacheeni Scott (Reference Toelken, Scott and Kroeber1981: 84–85) describe the over-abundant use of ąą in words at key moments in a Coyote narrative told by Yellowman. As they note, as a standalone form ąą is an affirmation interjection and thus its use adds a certain ironic texture to the key moments in the narrative that are not as they appear to the characters involved.

The seductiveness of this poem becomes most apparent in the next excerpt:

• AW: what does da'diłdił mean 38

• BM: da'diłdił is like throwing the stick repeatedly 39

• AW: throwing the stick repeatedly 40

• BM: yiits'a’ is its sound  41

• [5 sec pause]

• AW: but then yiits'a’ doesn't change that into the sound it's making 42

• BM: it could be both 43

• it could be dil dil 44

• or da'diłdił 45

• AW: okay 46

• BM: see diłdił 47

• dił by itself is blood 48

• you could say blood, blood 49

• AW: right 50

• BM: but in this case it's not blood, blood 51

• AW: right 52

• BM: it's a: 53

• da'diłdił would be just like leaping or prancing higher than the leap

• that would be diłdił 54

• so it's got three meanings dil dil 55

• it could be the sound 56

• it could be the leaping motion whatever that is 57

• yiits'a’ makes it its sound 58

In line 41, Mitchell states that “yiits'a’ is its sound”, and then there is a five-second pause. My distinct impression is that Mitchell was beginning to enjoy the realization that there was something going on in this poem. In line 42, I then make a statement – not a question here – that the verb of sounding does not change the meaning (my intonation contour is flat for this statement). Mitchell then indirectly corrects me with, “it could be both.” Indeed, after finding two possible ways to think about this line, Mitchell then adds a third way of thinking about it (“you could say blood, blood”). However, he quickly discards that view (“but in this case it's not blood, blood”). Yet in line 55, he then uses the resultative “so” and states, “so it's got three meanings dil dil” (interestingly enough, he pronounces the form with the voiced lateral and not the voiceless lateral). My understanding from the conversation was that the three meanings were: 1) “the sound” of the dice, 2) “the leaping motion” of the dice, and 3) “blood.” Finally, Mitchell's translation of the third line of the poem as “are stomping” suggests his own sensitivity to the sound evoked by this form (whether or not it is an ideophone or just sounds like one).

My point here is that ts'idil and tsidił and da'diłdił and da’ dil dil are potential puns in Navajo and that the final line (yiits'a’) encourages or seduces the listener or reader to hear or understand the poem as either composed of ideophones or that such ideophones – as evoked through the potential puns – might resonate around the poem (creating, that is, a layer of soundedness that envelopes the poem). To borrow Sylvia Wright's (Reference Wright1954) term, the mondegreens (the mishearing of the words of poem or song) here are motivated by phonological iconicity or homophony (words resembling other words through sound). Mondegreens become, in this view, one crucial window on the ways that languages facilitate imaginative acts (both within and across languages).

7. Stick dice resonances: blood, lightning and thunder

Having laid out something of the music of this poem and the seduction of ideophony, I now want to add some context (myth, in Friedrich's terms) about the stick game and suggest how this context might inform the resonances in this poem. David Aberle (Reference Aberle1942) has discussed the stick game and its associated mythology at length. Here is how Aberle (Reference Aberle1942: 144) describes the game:

Aberle then goes on to make a number of important points. First, as is often the case with Navajos, Aberle documents a number of origin stories for the stick game. Aberle (Reference Aberle1942: 153) notes that these are “equally” viable though different origin narratives. In my experience, attempts to ascertain from Navajo consultants the “correct” version of something were often met with the response of “it depends” (Webster Reference Webster2009: 208). Such issues often depend on context. Likewise, Aberle (Reference Aberle1942) noted that not all Navajos were willing to share information concerning the origins of the stick game with him. Aberle (Reference Aberle1942: 145) gives this revealing account, even if he is also dismissive of it:

Other reasons were also given:

I, too, worked during the summer with Mitchell and Translator 2 (paying them both for such work), but did not inquire too deeply about the stick game at that time (though some information was provided to me then and later).Footnote ²

In a passage reminiscent of the work of Barre Toelken (Reference Toelken, Swann and Krupat1987) concerning Navajo Coyote stories, Aberle (Reference Aberle1942: 146) notes the following:

Taking such comments seriously, in what follows I merely want to note a certain resonance between the forms in Jim's poem, the ways they were variously translated, and some features of the stick game. I leave it up to others to comment on the “meaning” of this poem.

The stick game is largely thought of as a women's game, though both men and women can play it. The association with women is related to an origin story that recognizes women as the first players of the game (Aberle Reference Aberle1942: 152) and/or that the important mythic figure Changing Woman took a hand in creating the game (Aberle Reference Aberle1942: 146–147). Young females at the time of their first period go through a rite of passage called Kinaaldá where they embody Changing Woman; this is one of the major rituals in Navajo society (see Frisbie Reference Frisbie1993, Schwarz Reference Schwarz1997a). That's point one.

Aberle (Reference Aberle1942) notes at numerous points that Navajos associate the stick game with lightning, rainbows and rain. For example, Aberle (Reference Aberle1942: 154) writes that “throwing stick-dice is called ‘like lightning,’ or ‘means a rainbow.’” Earlier, Aberle (Reference Aberle1942: 148) notes that “DS identifies the sticks with a different natural force: that of lightning.” This contrasted with “WH identifies the game with the world, and the sticks with rain” (Aberle Reference Aberle1942: 147). Finally, Aberle (Reference Aberle1942: 154) notes Navajos in the community of Ramah suggested “that playing the old games will bring rain and good crops.” That's point two.

Let me now reintroduce the three meanings that Mitchell suggested: 1) the sound of the dice (dil, dil); 2) “the leaping motion” of the dice; 3) “blood.” Working backwards, there is a resonance between the -dił in da'diłdił and tsidił and the lexical item dił ‘blood.’ That resonance between forms also resonates with the role of Changing Woman in the creation of the stick dice game and its association with women (especially as it relates to the Kinaaldá). Maureen Schwarz (Reference Schwarz1997a, Reference Schwarz1997b, Reference Schwarz2003) has discussed in detail the importance of blood (menstrual and other) in Navajo philosophy. Suffice it to say that the resonance of blood here is not a trivial matter.

As for the form da'diłdił, while it appears to have a reduplicated form here, that form is analyzable as di-, a classifier that Young and Morgan (Reference Young and Morgan1987: 25) suggest is associated with an “elongated object”, and -łdił “to cause it or them to move independently – a slender flexible object, as a rope or pair of objects, as shoes, stick dice” (Young and Morgan Reference Young and Morgan1987: 25). This is the verb stem -déél which Young and Morgan (Reference Young and Morgan1987: 129, 134) separate into two forms (though they leave open the possibility that the forms are related): -déél ₁ ‘abrupt movement of a ropelike object’ and -déél ₂ ‘jar, shake by concussion.’ The -ł- is a transitivity marker (causative-transitive). Young and Morgan (Reference Young and Morgan1992: 135) consider tsidił a noun derivative of -dééł ₂; though, at other times (see Young and Morgan Reference Young and Morgan1987: 25), they seem to link it with -déél ₁. In thinking about the throwing of stick dice (long objects thrown against a rock), they do seem to partake of both senses of -déél. In fact, following Margaret Field (Reference Field2009), one is tempted to see these forms as polysemous and not homophonous (see also Young and Morgan Reference Young and Morgan1987: 134). In looking at the many ways that -dééł ₁ is used, I note the following two examples from Young and Morgan (Reference Young and Morgan1992: 131):

1. (5) Atsiniltł’ish P-ił dah yizdéél

   P was transported by lightning

2. (6) Nááts’íílid P-ił dah yizdéel

   P was transported by a rainbow

The verb stem is potentially evocative of the movement of lightning and of rainbows (both of which the stick game is likened to). The “leaping motion,” as described by Mitchell, of the dice can evoke the movement of lightning and rainbows.

The ideophone dil, as we have seen from discussions by Mitchell and Translator 2, seems to suggest a rumbling, stomping or banging (clanging) sound. This ideophone is likely related to the more commonly attested form diil ‘deep rumbling sound, as that produced by the peal of thunder’ (Young and Morgan Reference Young and Morgan1987: 432; see also Landar Reference Landar1985: 490, Young and Morgan Reference Young and Morgan1992: 138). That the stick dice game is associated with lightning and that da'diłdił in the poem evoked for both Mitchell and Translator 2 a sound reminiscent of thunder is not, I think, too surprising. Indeed, ‘adildííl ‘rumble of thunder’ and ‘adiłdiłí ‘stick dice player’ are, according to some Navajo consultants, a potential pun. Da'diłdił, then, not only describes the movement of the dice (reminiscent of lightning), but also evokes the sound of thunder. The sounds the dice make, then, are evocative of the sound of thunder. Lightning and thunder are then intertwined here.

8. Seduction and Misunderstanding

Dingemanse (Reference Dingemanse2012) begins his piece on the coercive power of iconicity by noting that “humans are iconophiles: we love to connect form and meaning, and do so even when it doesn't seem warranted.” My goal in this article has been to look at the motivations behind the seductive power of iconicity and ideophony; that is, to show how, in this case, it is warranted. Jim's poetry often works because of the sounds that it evokes and convokes – that is, the way that key sounds in a poem evoke and convoke other words (Mitchell and Webster Reference Mitchell and Webster2011; Webster Reference Webster2013, Reference Webster2015a). Certainly, Mitchell's working through of the multiple resonances of da'diłdił is a case in point. It isn't just the literal meaning of this form that matters, but rather that through phonological iconicity (words resembling each other through sound) one can hear – and that's the verb to use – both dił ‘blood’ and da’ dil dil ‘rumble, rumble.’ The former can evoke Changing Woman and the latter the sound of the game being played and, also, thunder (which links back with lightning). Thus, the music of the poem can conjure complex mythic relations. The resonances work in Navajo. The language-particular sounds of this poem matter. A literal translation's deficiency is precisely in losing the sounds of Navajo; of other ways of hearing and hence imagining this poem.

Saad aheełt’éégo diits'a’, or punning, is an important feature of Jim's poetry, but it is also an important component of Navajo verbal life more generally (see Sapir Reference Sapir1932; Kluckhohn and Leighton Reference Kluckhohn and Leighton1962; Webster Reference Webster2010, Reference Webster2013). Punning can be found in curing chants (McAllester Reference McAllester1954: 80) and in everyday conversations (Webster Reference Webster2010). It is part of a broader aesthetic that seems predicated on an appreciation of both the precision of Navajo and on its ambiguity. To echo words of consultants, it both really describes something and, simultaneously, potentially sounds like something else (see also Peterson and Webster Reference Peterson and Webster2013). Navajo speakers of Navajo that I have discussed this with find the description of Navajo as both precise and simultaneously ambiguous an apt characterization. Though Mitchell once shortened that description to noting that Navajo is a “tricky language.” Jim's poems seem to work that way too.

Translator 2 takes this point a bit further. It appears that the yiits'a’ at the end of the poem seduced him into hearing the poem as full of ideophones. Mitchell too saw the use of yiits'a’ as encouraging one to hear line three as potentially ideophonic. But the reading of this poem by Translator 2 is absolutely intelligible based on potential puns in Navajo. Ts'idil does sound like tsidił and, as I noted above, da’ dil dil does sound like da'diłdił. Both puns were certainly available in Navajo. Because in Navajo, at least, such sound-imitative ideophones resonate with a variety of forms through punning and mondegreens, I am hard pressed to hear these forms as in any way simpler than other forms of ideophones. They are, or can be, deeply interwoven throughout and across languages (see Woodbury Reference Woodbury, Grenoble and Whaley1998 on interwoven-ness).

The example of a mondegreen, where the lines of a poem – the very words – are understood as not one thing, but another, also raises questions about the role of misunderstanding or not-understanding in a concern with linguistic relativity. To get at that discussion, I want to begin with Johannes Fabian's (Reference Fabian1995) work on the value of misunderstanding in language-centered ethnography. Fabian (Reference Fabian1995) argues for attending to misunderstanding and not-understanding, not as things to be explained away (solved, as it were), but for the ongoing dialectic nature of misunderstanding and not-understanding. He is concerned with forms of misunderstanding and not-understanding that range from phonology to morphology to semantics to syntax and pragmatics. In the example he discusses concerning sounds and phonology, he describes how he and later a language consultant transcribed the repeated refrain of a song from a play as tutabawina ‘we shall beat them.’ They identified the language as Shaba/Katanga Swahili and the genre of song as a “fighting-song,” perhaps associated with soccer matches (Fabian Reference Fabian1995: 42). Later, when Fabian inquired to the theater group that had performed the song, they replied with a letter which had the line as Tuta Mawila ‘march on’. The language was not Shaba/Katanga Swahili, but rather Kizela, and the song genre was a marching song, not a fighting song. It seems that tutabawina and Tuta Mawila are bivalent forms, sufficiently similar to be recognized as “the same.” This is, of course, a mondegreen (the mondegreen here is across languages and is very much like an interlingual pun). The context, a play largely in Swahili, seduced Fabian and his language consultant into hearing the song lyrics in the play as Swahili. Saying now that we have corrected that “misunderstanding” reduces the rightness of the bivalency – the simultaneity of the bivalency – to an either/or proposition. We have reduced misunderstandings to the “conceptual aura of mistake, error, failure, and falsity” (Fabian Reference Fabian1995: 48).

Such a practice – the explaining away of misunderstandings as mistakes or failures – denies the creative accomplishments of those that might hear Tuta Mawila as tutabawina or tsidił as ts'idil – the ways that misunderstandings are generative of understanding or generative of some kind of knowing or knowledge. For Fabian misunderstandings are not aberrant to languages in use, but rather deeply intertwined with it. Here Fabian quotes Wilhelm von Humboldt:

“Solving” the misunderstanding and treating it as a mistake; that is, telling you what was “really” meant, artificially attempts to halt the vibrations, ripples, or echoes of difference – the very stuff of language. Languages and their users are, to use an old phrase from Anthony Wallace (Reference Wallace1961: 28), “organizations of diversity” (see also Sherzer Reference Sherzer1970). A part of a “thick translation” of verbal art, then, should include annotations and exegeses on the ways that various forms are, in practice and when possible, misunderstood (on “thick translations” see Leavitt Reference Leavitt, O'Neil, Scoggin and Tuite2006, Woodbury Reference Woodbury and Austin2007).

Given that many of Jim's poems seem predicated on uncertainty and ambiguity, they highlight the multiple ways that a given utterance can be heard and understood. To say that Translator 2 made a mistake in his interpretation assumes there was a singular interpretation to be made in the first place – and that seems rather presumptuous. Navajos that I have worked with have evaluated ambiguity in Jim's poetry as a key component of what makes these “strong poems.” Strong poems are poems that provoke on-going contemplation and that do not force a singular interpretation. Echoing a general Navajo ethos that t’áá bí bee bóholnííh ‘it's up to him/her to decide’, Jim's poems are invitations to imaginative acts (see Webster Reference Webster2015b). Translator 2's translation, and then my later discussions with Mitchell based on that translation, allowed me to begin to think about how one might say that the stick game, as locally recognized traditional Navajos sometimes do say, is “like lightning,” in the ways that tsidił can be heard as ts'idil – the sound of thunder. There is something deeply “right” about his translation in that respect. Solving the “mistake” or “misunderstanding” ignores then both the potential pun and the knowledge that comes from thinking through that pun. There, in a nutshell, is a positive vision of linguistic relativity, the Boas, Sapir, Friedrich, and Sherzer vision: that particular languages, exemplified in and through poetry, facilitate ways of imagining. The sounds of the poem suggest resonances, but as Translator 3 shows, they do not determine them; they merely create potentials and possibilities for imaginative acts. Mondegreens, this mishearing of the sounds of poetry, are examples of the ways that languages can facilitate imaginative acts and suggest the creative possibilities of such imaginative acts.