Let’s pretend (and pretend we must) that I am a physicist and I open my prepared remarks before an audience like this: I would like this evening to say a few words about string theory, so that you might understand the basics of this complicated but amazing theory about the universe. Writing well involves not only questions of grammar, but also rhetorical choices, choices of style. In this opening sentence, where one decides to position the phrase this evening will change the character of the statement ever so slightly.

We should understand first that the phrase this evening is an adverb, specifically a temporal adverb, because it will be saying something about the time of whatever idea it is ultimately positioned to qualify. We should understand, too, that the temporal adverb is only one of a number of different kinds, each limiting in a different logical way the element it modifies. Adverbs of place, for example, have to do with location: I spoke to him in the lobby. Adverbs of manner concern the way in which something is done: that dog barked furiously at me. Adverbs of degree mark the magnitude of an action: they drove very fast down a country road. Such is the complexity (or probably better, the intricacy) of human awareness that grammarians have found numerous classes of adverbs to distinguish among; a practical number to recognize when revising our work, however, is four: time, place, manner, and degree.

Adverbs most often modify verbs (though they can modify adjectives and other adverbs as well), and where the adverbial phrase this evening stands now in our example sentence makes it unclear whether it is working with the preceding verb phrase, would like, or with the following infinitive, to say. In theory, which means here as a general rule, adverbs are governed by the law of proximity: they stand as close as they can to the word they modify. Whether that will mean before or after or between the modified elements will depend upon the kind of adverb it is. Adverbs of time, such as the phrase this evening, can be moved around quite liberally to rich and subtle effect.

If standing in this ambiguous position produces more uncertainty than I want to suggest, I could move this temporal adverbial phrase to the first position of the sentence: This evening I would like to say…. This, too, is an acceptable choice, but as a writer (or speaker) I should recognize that in this primary position, I am giving a decided emphasis to the idea of time: this evening, this particular occasion. The first and final positions of a sentence are always the most emphatic, and so if I decide that this emphasis is misplaced, I could reposition the phrase to stand immediately after the infinitive: I would like to say this evening…. That, however, might put an equally inappropriate emphasis on saying, as if to suggest I were about to assert something in opposition to something else, almost as if to be defending myself. Or I could decide to finalize the entire first clause with the adverb, establishing a more conclusive tone: I would like to say a few words about string theory this evening….

The point to notice here, then, is that so much depends on the right placing of the elements in a sentence. Not to hear the differences between these many rhetorical choices is to miss the chance both to create and to enjoy more of what lies beneath a simple-seeming thought. We’re often too ready to settle for less. My neighbor’s son, who is four years old, assured me the other day that he doesn’t need to study any more math when he begins first grade this fall because he “already knows math.” Preschool, it seems, was enough. There’s a lesson there, I thought: we all run the real risk of shortchanging ourselves when we believe there’s nothing more to discover. Though we inevitably come to realize how very wide and deep and rich is the world we find ourselves in, that expansive wealth itself can dull us early on into a routine of diminished expectations. Watching closely, as we must in the study and practice of writing, will help us keep our expectations high and wait for surprises everywhere—even in first-grade math.

***

Poetry has been described as a “concentrated language” whose words are freighted with meaning which the reader must take on and carry back to the idea they arose from. That phrase “concentrated language” belongs to the distinguished literary critic R. P. Blackmur, and he goes on to say that “poems remain obscure until the reader takes out what the poet puts in.”

Prose writers, unlike the poets, cannot expect their readers to meditate so deeply on the construction of their language. The contract they hold with their readers captions an entirely different set of mutual obligations. Clear prose must be clear first on the surface, and what may be found more deeply within the grammar should still not be necessary to understand the immediate significance of the sentences, what Blackmur calls “the bare indicative statement of experience.”

Take, for example, this profoundly prosaic sentence: That’s the best cake I ever ate. The surface meaning is clear enough quickly enough. The notions of a cake, of someone called I, of the action of having eaten are all so obvious that we cannot call this poetry in any real sense, because all of those ideas are straightforward, suggesting little beyond themselves. They are the facts they are, not loaded down by hints of deeper meaning. Still, the grammar of prose may be fraught with suggestions of a richer understanding if we take the time to ask questions of style and construction.

One such question might be this: what is the difference between the best cake I ever ate and the best cake I have ever eaten? The material difference lies with the verbs. Ate is the simple past tense of the infinitive to eat, and have eaten is its present perfect tense. The subjects of the two sentences (I) remain the same, as do the direct objects (the best cake) and the adverb ever. But it’s that last grammatical role, the adverb, which changes the cast of the two sentences. Understanding how is the contractual obligation of both the writer and reader.

Adverbs partially change the meaning of a verb or adjective or other adverb. The traditional terminology for this change is modify, and in the Latin etymology of that term we can see the function of an adverb more exactly: making or changing (-fy) the way (modus) in which a verb (or adjective or other adverb) means what it means. The adverb ever represents the idea of at any time, and in each of these two sentences, it is modifying the verb. The phrase at any time is a temporal idea, and a temporal adverb is just one of a number of different kinds: there are also adverbs of manner (well), adverbs of degree (very), adverbs of cause (why), and adverbs of result (therefore). So how does adding the idea of time to the verbs in these two sentences change the implications of each?

The simple past tense (ate) is meant to take a snapshot of a moment, to represent a slice of time in a time now gone. The present perfect tense (have eaten) refers not only to a single past moment, but to a stretch of the past whose effects are felt to continue almost to the present. The simple past makes no more exact a claim about a past moment than that something happened in the past somewhere; the present perfect suggests the possibility of an immediate past, one so recent, in fact, that its effects might still be felt. I can say I ate a piece of cake a month ago, but I can’t say I have eaten one a month ago. That past is now long gone.

So, to add the adverb ever to the simple past ate in the first example is to contort the purpose of that tense by making it try to extend its meaning at any time, even to the very recent past. But that is done better by the present perfect tense in the second sentence, which already has in its design the very purpose of including what might have happened only a moment ago. The character of a sentence depends in good measure on the perceptual character of the writer, but for that to mean something to us, we readers must bring our own awareness of how the language works if we want to gain the most we can from a writer’s words.

***

Let’s try to wrestle down an idea about writing which is, I’ll admit from the start, a bit slippery. It begins with an assumption so simple that its implications for composing clear prose can escape us: we think about things. When we have an idea, it’s an idea about something. When we speak to someone, we’re speaking about something. When we try to understand what we’re hearing, we’re listening for something. Our mental activity, what is sometimes called our waking consciousness, is transitive; it aims at objects, at things, whether material objects in the world or mental objects in our mind.

Writing is often as much an effort of discovery as it is one of just transcribing the ideas which pass before our mind’s eye. We’re looking for something when we begin to compose, and that’s why an initial draft can often end up vague or crumbly: there’s nothing there to sink one’s teeth into; it has no body, no density. We often succeed in throwing some ingredients into a first draft, but they don’t cohere and take on a certain shape because what’s being written about is not even clear to the writer. It’s all really just exploratory. There might be an interesting, even an unusual and compelling idea there, but the writer has not seen it distinctly enough to say something specific about it. Writing well is more than adumbrating.

We can move from a dissatisfying draft to a redoubtable final revision, though, if we take a moment to inspect the words we’ve put together in a first try. We advance by improving what needs improving, not by deleting the entire document in a fit of pique, nor by (and one can see this all the time) cutting and pasting pieces of this sentence into that one, or blocking and moving entire paragraphs from here to there and back around again, hoping beyond hope that it will all start meaning something. That is to write by assembly, not by organic growth. And even though the pedagogical tradition of writing has enshrined the word mechanics in designating the more superficial matters of orthography and formatting, to properly inspect our early work does mean to analyze it, if not mechanically, then closely.

Here’s an example I have modeled off a description I found online recently for a program of study at a major university: The mission of our department is anchored in the belief that a variety of experiences and vocabularies are deeply essential in order to access an appreciation of other cultures and practices. When we encounter a piece of writing like this, examples of which are legion, the only way to proceed is to call up our conscious, transitive mind and ask what the writer is talking about. I italicize about in order to emphasize the fact that the writer is under an obligation to name something, and more often than not to name only those things which, taken and woven together, advance to a clear conclusion.

This example is a perfectly fine first draft; it is not a polished product fit for public presentation. We reach the latter through the former, though, and so a close inspection of the draft will bring our attention first to the opening ten-word clause. There the writer is speaking about a mission, a department, the action of anchoring, and a belief, all sewn together into a metaphor which is no more effective in the context than simply saying we believe. The first half of the following subordinate clause speaks of variety, experiences, vocabularies as things which are deeply essential, but the nouns are so wildly vague (pluralizing the last two is merely an ill-fated attempt to concretize the singular abstractions) that we are hard-pressed to see in any mental light, however bright, what exactly is so essential. And the second half of this same clause, in order to access an appreciation of other cultures and practices, grandiloquently proclaims that the purpose of the department’s mission is to access the abstraction appreciation of apparently multiple unnamed societies.

What, then, is left after such searching for what the writer is talking about? Merely this: We believe that it is essential to appreciate other cultures. Not much, but it is a start, and all one would have to do now would be to think hard, which means to think specifically, about the meaning of the pronoun it. What is it, what, what, exactly, that is so essential in appreciating other cultures? Which other cultures? And what does to appreciate something mean precisely in this context? When the writer can answer those determinate questions, the way will be open to something we readers will be able to think about. And then we’ll all be on our way together toward a mutual understanding of some degree, on which civilization (and that word, certainly now, is not too grand) depends.

***

It is largely true, I think, that most of us are comma happy—happy, that is, to bisect and trisect our sentences wherever we might hear ourselves taking a breath, either subliminally or aloud. Sometimes that heuristic works and sometimes, with real consequences, it doesn’t. But if we don’t use our ear, what other way is there to determine whether a comma is needed somewhere or not?

Let’s take this sentence I’ve modeled grammatically on one I saw last week in a major publication: My bank confirmed yesterday that I needed to provide further details about my stock investments, which performed poorly last year, despite their impressive ten-year average. Some believe (and I am one of them) that we are too ready to cut up our sentences because we fear producing a heavy, dense prose which will tax the reader’s attention. A comma here and there, we suppose, might thin things out and increase our chances that the reader will stay with us to the end.

That propensity to follow a hunch, however, is a mistake, because a jagged sentence which flails like a ragged flag in the wind is just as likely to weaken the reader’s perseverance by drawing attention from one concentrated center to a multiplicity of nubs in phrases and clauses across the sentence. Our example illustrates a fairly long complex sentence of twenty-seven words, with one independent clause governing two subsequent subordinate clauses. Because the main verb confirmed signifies the idea of saying or thinking, it triggers what is called indirect statement, and in English this often necessitates a noun clause beginning with the subordinating conjunction that: that I needed to provide further details about my stock investments. This first subordinate clause holds within itself another subordinate clause, which performed poorly last year, despite their impressive ten-year average, and it is the comma in this second dependent statement that is flapping in the wind here.

The word despite, meaning notwithstanding or in defiance of, is a preposition, and here its object is the noun average, or to include the three adjectives which modify that noun, their impressive ten-year average. The writer seems to have felt, consciously or not, that a comma was necessary to set off this prepositional phrase in order to make the sentence easier to comprehend by sectioning it; but this resulting prepositional limb of the sentence individualizes what is logically necessary to the preceding part of that same subordinate clause, and that logical isolation is sufficient cause to judge against it. The prepositional phrase (not just the preposition alone) works as an adverb, and the verb it is logically modifying is performed, the verb of the subordinate clause in which the phrase appears. To separate that phrase with a comma from the rest of its integral clause is to shake the logic of the sentence, not to break it outright, but to shudder it enough to make the reader wonder for a second too long about the connection of the two sections. And that logical wondering is just what clear expository prose has to avoid.

What makes all of this difficult is the fact that writing is not mathematics. Yes, there are, as we’ve just seen, objectively right and wrong answers in matters of punctuation, but then there are also decisions which, unlike the rigor of numbers, can only be made with reference to effect, what is called rhetorical punctuation. If, for example, that concluding prepositional phrase in our example included words that suggested a significant ellipsis, then a comma would be justifiable: and this despite their impressive ten-year average. This ends up constituting an entirely separate, though truncated, clause (and this decline in performance was in despite of their impressive ten-year average), and its emotional impact, the sustained surprise the current loss of value evokes in the writer, would be well served with the comma.

In writing, as in all the arts, even the political, rules are there to temper license and result in freedom. A balanced tension, always difficult to achieve, is still the real goal.

***

In the best of all worlds, our study of grammar and sentence structure should lead us out of the textbooks and into the precincts of literature and philosophy. All those rules and technicalities are right and good, but they are ultimately instrumental, there to serve a purpose higher than themselves. Without them, of course, we flounder; with them, however, we find solid footing from which to reflect on what a text may mean and imply.

A student of mine illustrated this close relationship between structure and meaning recently when he asked about the punctuation of this opening sentence of the chapter entitled “Baker Farm” in Thoreau’s Walden. This quotation of forty words represents only about one quarter of Thoreau’s entire sentence (it covers some twenty lines of printed text), which never once loses its poise and balance. Should there be that comma after groves, my student wanted to know:

Sometimes I rambled to pine groves, standing like temples, or like fleets at sea, full-rigged, with wavy boughs, and rippling with light, so soft and green and shady that the Druids would have forsaken their oaks to worship in them….

All the marks of punctuation in English, from comma through semicolon to period, are meant to pace the run of ideas moving across the reader’s attention. Just as we know that an aircraft cannot lift off the runway unless it is moving, so a sentence cannot convey meaning unless the reader’s eye is setting off through its words and phrases. Unlike the pure physics which govern winged flight, however, syntactical flights of fancy or fact must take regard not only of logic, but of emotive impressions as well. This is why the study of writing speaks of both logical and rhetorical punctuation, the former concerned to make clear what is being said and the latter interested in how that meaning is touching the reader.

The comma after groves dissociates that noun from the adjacent participle standing, which as a participle is modifying that noun in order to make the comparison of the groves, or more properly the pine groves, to temples. The comma frees the participle from that immediate reference, running the risk that the reader will think that the subject I is instead what is standing like temples. Were this passage to be brought up on charges only under the statutes of logical punctuation, there would be little hope beyond throwing it all on the mercy of the court: the participial phrase standing like temples is certainly meant to name a defining characteristic of the pine grove, even there in the predicate position, and such restrictive modifiers are never to be set off by commas.

But is the phrase standing like temples in fact certainly meant to define the pine groves? What is certainly certain is that Thoreau thought surely he could rely on the reader to follow him past the laws of punctuation and still logically equate those temples with the plural pine groves and not the singular I. The letter, after all, killeth. By isolating the otherwise restrictive phrase standing like temples and, for that matter, its alternative comparative phrase or like fleets at sea, Thoreau is sailing under the flag of rhetorical punctuation, whose authority will allow bending—but not breaking—the steel-cold rules of logical punctuation. It’s all a delicate matter of perception and expression.

As the English novelist and poet John Cowper Powys said a century ago now in The Meaning of Culture, “To philosophize is not to read philosophy; it is to feel philosophy.” We can misunderstand that and believe that the right way to do things is nothing more than a matter of how we personally want to do things. But when we do not forget the objectivity of the world we are writing amidst, we can understand Powys’s complementary thought:

The art of self-culture begins with a deeper awareness, borne in upon us either by some sharp emotional shock or little by little like an insidious rarefied air, of the marvel of our being alive at all; alive in a world as startling and mysterious, as lovely and horrible, as the one we live in.

I think so. And to write with that in mind is to hope to write as masterfully as Thoreau did.

***