I give you a new definition of a sentence:
A sentence is a sound in itself on which other sounds called words may be strung.
You may string words together without a sentence sound to string them on just as you may tie clothes together by the sleeve and stretch them without a clothes line between two trees, but—it is bad for the clothes.
Frost's definition is excellent for Frost. What poses as a meaningful definition--a logical definition of a sentence--ends up as a rhetorical introduction for a critique of other verse. In other words, the significance of his definition is not the so-called definition but the claim that all else is "bad for the clothes."
O, come to the defense of Frost if you must. I will say now: I read his poems. I'll say now: I have a genuine interest and admiration for Frost's work. I'll make it clear: a working knowledge and continuing learning of prosody is essential to a well-rounded education of poetry, of poetics. The problem is his explicit prohibition and the circularity in his reasoning. Quite simply, the definition assumes that the sentence structure exists prior to the line we recognize as a sentence after it is written: uses the word SOUND to anchor the sentence object as a SOUND object to address poetics and bracket linguistics. The problem is Frost "sounds" like Kant; the sound itself is analagous to the thing itself. And I agree with Nietzsche's critique of the thing itself.
In the case of poetics and sound itself, a reliance on the existence of such is actually a deferrence to a higher authority and a continually willful or voluntary surrender to that authority. Such an artists surrenders either fully or partially the will to create. And what he or she doesn't fully surrender is guiltily enjoyed and deformed. This produces within the artist, in this case poet, a slavish reliance on form without personal investment in cultivating new forms--in fact is prohibited from accessing such investment. To digress: we may wish to link such an attitude towards the new--locating it in tradition, in renewal--to the falling out of poetry, not from the market, but from the people. Why read poetry? is the question from the public to the poet. Why read poetry if it sounds the same, if it is what IT is? If it is not capable of innovation (renovation)? To return: this perspective promotes an attachment to a market of ideas constructed to cultivate ideological relationships to the exchange of poetic objects as economic objects. This process gets us counting while we should be emoting.
And, it provides authorities like Frost, an infamous grump (and that is kind,) a position of power that has the flavor of white masculine power structures. Because the sound itself: is it English or Russian that this sound comes from? Is it Teutonic or Mandarin? Who deserves to choose which sound in human language is more primary? Or which cultural factors permit specific artists to continue to construct such definitions on behalf of all other authors? I should ask many questions in this examination, questions the artist resists at best or disregards as beneath him. At any rate, I can think of many historically celebrated poems that do not illustrate Frost's "new" definition.
And what is "new" about it? Is it because is has the sound of the modern?
If a sentence is to be defined as A SOUND IN ITSELF (an object we'll call X) upon which WORDS (also sounds in themselves, but secondary objects while finite not worth assigning a single variable) are strung, then the sentence itself as an empty structure is prior to the sounding of the sentence. Think of Thoreau's discussion of the appearance of the loon and the call of the loon at Walden Pond. The relationship of an object's appearance is always troubled by the sounding of the object itself. The sentence is not necessarily the sound that must be there prior for the utterance of the thought to appear properly. Sometimes the sound is wrong when it is correct. In such cases, literary artists above all other speakers and users of language have the ability to break free from the self-sounding others and reconstruct the sentence itself in order to actually write a NEW SENTENCE. At any rate, If the above rephrasing of Frost's "new definition" weren't accurate, then his new definition would be patently silly. In other words, according to his defintiion, the structure for X must exist in each appearance or case before the sound of X and the sounds of all the words are secondary to the predetermined sound of X that as a fact must contain words to be said to have been uttered.
Frost implies thoughts--what bubbles up from being and erupts as words, phrases, then clauses--conform to language. And I think most of us would agree with this. To some extent we must cater to the capability of language to utter thought. We do say, for example, "I am searching for just the right words." But wait...according to Frost, we must know/recognize the right sentence that has a determined sound that restricts the use of words--syntax, order, length of breath, tone, all of our technique really--to sounding out the sentence properly. Or we ruin the clothing (that language gives thought. Of course, I agree with Wittgenstein that language is not primarily designed to expose the form of thought.) Frost's definition can serve no practical purpose other than to distinguish correct and incorrect presentations of language in verse because it simply refuses to address language as a whole. In other words, his new definition is not of the sentence but of style. And as such his is more of an attempt to claim that taste--what sounds sound proper (good and useful) to the ear--is predetermined and not up for debate.
The sentence, for Frost, defines the sentence. It's like tuning the orchestra--A must sound like A to be A, so all artists play a line of A in chorus and that line sounds like A or it isn't A. Works fine for tuning, but certainly has little to do with the sounds produced in chorus during the performance of a composition. In fact, many great composers have experimented with exactly this problem in the production of sound. I think of Terry Riley's In C immediately only because I was listening to it this morning. I think of Jazz improvisation as well. I think of what Coltrane does with breath; I think of what Olson does with breath. All violate the prohibitions of such a restrictive definition. Or, by his definition, they are bad for the clothes. Of course, this is only the case if you believe that language is designed to perform such a specialized task primarily and all others secondarily. That his definition is the telos of the sentence and all others are degraded forms. Not in my language.
The definition is useful, however. In other words, it is not THE definition of the sentence. In this way, it may have been A new defintion, or one new sentence. His is a good argument for the use of templates through which to temper our expression in verse (or prose.) Regardless, writer's have good reasons to refuse templates, as many if not more as there are reasons to use them.
I hadn't time to really revise this much before posting it about 90 minutes ago, so this version is in fact revised with a few qualifications. And I now add what I was going to end with before I had to leave my office to teach: I might be more inclined to agree with Frost if he had written something like, "A sentence is a line on which words are strung." Drop that sounds business.
A claim, then, to gather similarities from above:
Sounds are not lines.
I might add, though I haven't considered at length, that sounds are not structured like lines. Lines have necessary obligations that sounds ignore. It may be that sounds are asked to oblige but simply refuse regardless of our needs. As language can escape us in our moments of need to communicate positively private experience, sounds can betray us in our moments of need to communicate appropriately public experience.
Or something like that.