On tenor and vehicle

I wonder whether our discussions of metaphor have become a bit sloppy. When we find metaphor being treated merely like logic, as in the work of Paul de Man, something has gone wrong. We then take the metaphor as a unit, and, if we don't completely overlook its internal complexities (treating it as rhetoric in the abstract), open the door for all sorts of mistakes in actually determining how the figure is working (accepting any grammatical work at all as metaphoric). This probably extends to how de Man treats all figures, but I'd like to stick to metaphor, because so many others tend to treat it in his fashion.

Of course, I can say that things have become sloppy because we literary critics once (in our readings) had a very precise way of talking about the internal aspects of metaphor: through the distinction between tenor and vehicle. And while this famous formulation of I.A. Richards presented in The Philosophy of Rhetoric ends up being put to ends much too formalist and too psychological, one has to recognize that it seems sophisticated in comparison to how de Man talks about the figure (for example, in his analysis of passages of Proust in Allegories of Reading, or even his strained reading of the word "translucent" in the famous essay "The Rhetoric of Temporality").

So what are the tenor and vehicle? They are the elements of the metaphor, its two components. This is the first and most fundamental point which must be understood: the metaphor is the name for the copresence of the two elements in the form of the two sets of ideas being related, and never is reducible to either one. For it is still (still!) common to say that the figure itself is the second element, the vehicle.

Let me use an example: when I say that my love is a red red rose (to modify Burns a bit), taking the metaphor as the vehicle would mean saying that the red rose is the only thing in the sentence that is metaphoric. Richards' real innovation is to understand that this requires my love to be unmetaphoric, or the literal, plain meaning which the rose simply embellishes.

Now, this might seem exactly right with the use of a simile: what is the rose here doing except embellishing, coloring the woman, my love? But extend it to any more complicated (and normal) form of a metaphor and you immediately get confusion: when I talk about how the morning sun kisses the mountainside (as Shakespeare does in his 33rd sonnet), or even how my dessert is kissed with honey, we're not using using kiss to embellish some plain meaning.

Yes, we could explain what we mean otherwise--a small amount of honey is delicately added to my dessert--and thereby act as if the plain meaning is being added to by some fancy-schmancy language. But to do this, look at what is necessary: an expansion of the vehicle, not explaining the plain meaning which this vehicle supposedly embellishes, the tenor. In other words, we have moved over to what we're calling metaphor itself (and what is really the vehicle) in order to retain some notion that there is a sense (a tenor) which is unmetaphoric. We have made precisely what we called metaphoric (and is really the vehicle) into something other than a metaphor in order to explain how it (the vehicle) is metaphoric (as a mere embellishment)!

It makes sense, then, to change this situation and say that what we're calling metaphoric is only one part of the metaphor: the vehicle. For once we do this, in explaining the vehicle we can still say that we are explaining a metaphor--we don't have to literalize what is metaphoric in order to prove its metaphoricity. This is the brilliance of Richards' distinction: it recognizes the way we are looking at the problem in the process of our reading and, by giving us a clearer terminology, actually allows us to continue what we're doing, but much more precisely. He calls this a process of translating our skill into observation and theory, by which he means making our skill in reading and deciphering the problem explicit so as to refine it. And this is an alternative to what people still do when unable to articulate how their precise readerly observations come from more than just mere insight: fetishize the skill, turn it into the mysteious capacity that only a few distinguished individuals possess. Richards, yes, came up with the phrase "close reading," but he uses it in the exact opposite way of those who talk about Derrida, for example, as an "extremely close reader."

If we say that the metaphor involves the copresence of the tenor and the vehicle, then, we can begin to investigate the different ways this copresence takes place. We can talk about the different interactions of the tenor and vehicle, and see that embellishment is only one possible relation between the two. We can even begin to make sense of what at first appear like limit cases, because, with a recognition of the vehicle as vehicle, the tenor becomes much more expansive than "plain meaning." No longer identified with such meaning, it can become even something as generally unmeaningful as the excuse for merely using a vehicle.

In Shelley (I'm thinking of Epipsychidion especially) we have a lot of the this latter phenomenon, along with the layering of metaphors (the use of a vehicle as a tenor for another vehicle) which Richards' vocabulary also allows us to explain. However, Richards' example is John Denham's description of the Thames in what is usually called the first locodescriptive poem, "Cooper's Hill" (1642)--an example which Samuel Johnson, in his Life of the poet, praises:

O could I flow like thee, and make thy stream
My great exemplar as it is my theme!
Though deep, yet clear; though gentle, yet not dull;
Strong without rage; without o'erflowing, full.

No doubt Johnson had other reasons for liking this (its intense reliance on parallelism, that great mechanism of Johnsonian thinking), but he says that it is good because "the particulars of the resemblance are so perspicaciously collected." We'll come back to this in a moment, but for now, let us see how Richards explains what is really the tenuousness of the relationship between the tenor ("I," or, as Richards says, the poet's mind) and the vehicle (the river) in the last lines and its development of what at first appears to be "a limit case" of metaphor:

The more carefully and attentively we go over the senses and implications of deep, clear, gentle, strong and full as they apply to a stream and to a mind [...] the more will the vehicle, the river, come to seem an excuse for saying about the mind something which could not be said about the river. Take deep. Its main implications as regards a river are, "not easily crossed, dangerous, navigable, and suitable for swimming, perhaps." As applied to a mind, it suggests "mysterious, a lot going on, rich in knowledge and power, not easily accounted for, acting from serious and important reasons." What the lines say of the mind is something that does not come from the river. But the river is not a mere excuse, or a decoration only, a gilding of the moral pill. The vehicle is still controlling the mode in which the tenor forms. That appears at once if we try replacing the river with, say, a cup of tea!
-The Philosophy of Rhetoric, 122-23

Richards then quotes again the last two lines just to show you:

Though deep, yet clear; though gentle, yet not dull;
Strong without rage; without o'erflowing, full.

If the vehicle used here was tea, and not a river, the lines obviously become ridiculous: while a tea can be strong, and while a cup of tea can deep and full (without overflowing, or rather spilling!), no one would say it is "strong without rage" unless they were using the vehicle to give us a better sense of how a mind works. Thus, we can see that a river is a better vehicle than a cup of tea. That is, the vehicle is still related to the tenor, developing it, even if it has come so far away from the tenor that "what the lines say of the mind is something that does not come from the river." The tenor, then, is no longer a sort of pure meaning which the vehicle modifies, but something which, precisely "forms" as the vehicle "controls" it.

Thus the relationship between the tenor and vehicle can modulate to accommodate, on the one hand, a situation where the vehicle merely develops what is already pretty apparent in the tenor, and, at the other, a situation (this one here) where the vehicle takes up senses that do not resemble the tenor at all but which still lend it a more powerful meaning.

And with this we can turn back to Johnson. The vehicle we see, doesn't have to have any "perspicaciously" achieved "resemblance" to the tenor at all: Richards takes up this example to show that what Johnson praises about the relationship between tenor and vehicle ultimately would have to make the vehicle still embellish the tenor, and restrict the tenor only to something like the plain meaning underlying the vehicle. Rather, resemblance is only one way in which the tenor and vehicle can be achieved, just as embellishment is only one way in which the tenor and vehicle can relate. More often, there will be a process where the vehicle simply modifies or qualifies the tenor in the way that the vehicle of the river does here.

To describe this relation more adequately, Richards begins to supplement his original schema.  At present, we have simply the following:

metaphor:

tenor + vehicle

Now, Richards introduces another concept, the "ground" of the metaphor. Most accounts of Richards' distinction leave this out, though it only describes more clearly what we have been talking about concerning the resemblance of the vehicle to the tenor. The ground of the metaphor is merely the presence of a tenor-vehicle relation, which is most "solid" (or the least "recondite," as Richards will say, 117) in the form of resemblance. Thus, where we have a vehicle that relates to the tenor in the form of resemblance, the metaphor has a ground. And where we have a metaphor where the vehicle is controlling the mode in which the tenor forms, but not by bringing out something that resembles the tenor, the metaphor has less ground (i.e. the metaphor is more ungrounded). A particularly common form of metaphor where there is less ground is found in instances where we call someone a "pig," for example, if we are disgusted by them. There is no relationship of resemblance, but there is a sort of commonality between the tenor and vehicle here which ultimately is the function of the ground: it is the common characteristic that sets up a relation between tenor and vehicle. However, this commonality is not in directly mimetic terms, so there seems to be less ground. Instead, there is only a commonality in terms of how we feel about the person and how we feel about pigs.

So, to reiterate, where the relation of the tenor to the vehicle is one of resemblance, there is ground. Where there is less of this type of relation, there is less ground--though the ground still hasn't disappeared. However, we see now that in neither case are we talking about the relation of the tenor and vehicle solely in terms of each other, so we can easily admit what we previously had to talk about as "limit cases" as what they really are: basically normal metaphors. The "limit cases" are normalized, not in order to disturb our assumption that we talk in plain language most of the time (though this view of metaphor will allow us to see "most sentences in free or fluid discourse turn out to be metaphoric," and that "literal language is rare," 120), but to be truer to the way metaphor is used, which is often not at all like a simile--that is, as an embellishment. So, let's modify the schema, incorporating Richards' changes:

metaphor:

tenor ↔ vehicle
------ground------

With this fully developed model, Richards is in fact able to tackle the famous table leg. Leg of a table is generally recognized to be not really a good, pure metaphor, but a catachresis. However, as the catachresis is usually defined as a failed metaphor, we have to actually explain on what basis we would call the leg of a table a leg. No one, however, ever explains this, except by saying that the metaphor is a dead one (which doesn't tell us much)... no doubt because it is simply easier to say why the leg fails as a metaphor (tables aren't human!). Richards, however, with his more developed schema, is able to explain why it is a metaphor quite clearly:

Let me begin now with the simplest, most familiar case of verbal metaphor--the leg of a table for example. We call it dead but it comes to life very readily. Now how does it differ from a plain or literal use of the word, in the leg of a horse, say? The obvious difference is that the leg of a table has only some of the characteristics of the leg of the horse. A table does not walk with its legs; they only hold it up and so on. In such a case we call the common characteristics the ground of the metaphor. Here we can easily find the ground, but very often we cannot. A metaphor may work admirably without oru being able with any confidence to say how it works...
-The Philosophy of Rhetoric, 117

Thus while the leg of the table has less ground than, say, the example of the river, there is still a commonality, still some form of ground between a tenor and vehicle, and so it is indeed metaphoric. For--as we said above concerning the normalization of "limit cases"--what Richards is trying to admit through this notion of ground is the case where,

the peculiar modification of the tenor which the vehicle brings about is even more the work of [the tenor and vehicle's] unlikenesses than of their likenesses.
-The Philosophy of Rhetoric, 127.

So while a leg only shares one real commonality as a vehicle (in the action of holding up), it is still part of a metaphor and should be recognized as such, precisely because there could be cases where the lack of resemblance will (unlike here) make a better metaphor. These instances aren't an instance of the paradoxical good catachresis, but simply metaphors.

This is also because is on this new basis of ground that Richards distinguishes between metaphoric language and non-metaphoric language: we can say that what goes beyond our "limit cases"--a metaphor completely without ground and a metaphor where there is too much ground such that the vehicle and tenor are identical--will not be a metaphor. Why? Because no relation will be there between a tenor and a vehicle, such that we can distinguish between them and relate them:

If we cannot distinguish tenor from vehicle then we may provisionally take the word to be literal; if we can distinguish at least two co-operating uses, then we have metaphor.
-The Philosophy of Rhetoric, 119

I put this in terms of ground because while Richards is emphasizing here how one idea described by the same idea is not metaphoric, we can also stress (perhaps against Richards' tendency to psychologize) how the opposite case also holds: when we relate two things without any commonality whatsoever, this is a literal relation.

One final point. I stressed the presence of ground, and the possibility of the metaphor that still works quite well when it is less grounded, because Richards introduces one more concept immediately after talking about the table leg--that of the shift between tenor and vehicle across the ground:

...A table does not walk with its legs; they only hold it up and so on. In such a case we call the common characteristics the ground of the metaphor. Here [in the case of the table leg] we can easily find the ground, but very often we cannot. A metaphor may work admirably without our being able with any confidence to say how it works or what is the ground of the shift.
-The Philosophy of Rhetoric, 117

Now I call the "shift" another concept, but what is really the case--and should be clear from the entire exposition here, and which, even more strikingly than the lack of talk about ground, I have also never seen elaborated--is that all of the terms which Richards introduces to explain metaphor are also metaphors: tenor, vehicle, ground, and now shift. I bring this up now because this metaphor of the shift allows us to describe metaphor completely metaphorically, but also more accurately, because all the other terms are able to give a fuller sense to as well as explain metaphor. For indeed, when we only use "tenor" and "vehicle," which is what happens when (rarely, as we noted) such terms are brought up, we do only the latter--and such explanation tends to make the terms into what they were meant to replace. We can complete our diagram, then, as follows:

metaphor:

tenor             
shift→
             |vehicle|
------ground------

Richards is comparing metaphor itself, in other words, to something like a car, such that we don't even have to talk abstractly about the relation of the tenor to the vehicle. Rather, the vehicle brings the tenor along, and shifts it across a ground. The tenor gets in the vehicle, and the vehicle shifts it. Thus we see why Richards chose "vehicle" and talked about how the vehicle modifies the process of the tenor's formation: the vehicle takes over the movement of the tenor. Furthermore, because we stressed the possibility of the attitudinal nature of the commonality or ground, we see also why Richards chose "tenor" for the name of the first element in the metaphor: the tenor is like tone--it isn't what the message means, but it is what sets the message in a particular direction (tone being the attitude of the speaker to the listener), allowing it to develop (here) by other means (the vehicle, with which it can share a common attitudinal relationship). In other words it is as much of a mistake to see the meaning of the metaphor within its first element (the tenor) as it is to see the meaning of a word solely in how it is delivered, in its tone: how the message is delivered isn't the same thing as its meaning, and by emphasizing how what we normally think of as the meaning could really only be attitudinal, our mode of trying to talk about the metaphor in terms of its literal beginnings is hindered. Nevertheless, while tone or how the message is delivered isn't the same thing as meaning, it certainly allows the message to develop: the metaphor is where such a message in a word is developing in a complex way through another word.

In short, all the metaphors explain and modify each other in this new structure, which the final diagram above can only somewhat adequately explain (to be more accurate, we would have to act more as if the tenor is moving over the ground by way of the vehicle and through the shift--thus the shift should also be represented more as the shift of the tenor and vehicle over the ground). Nevertheless, if you begin to use all these metaphors, you have a sense for how complicated and rich the internal complexities of metaphor become.