Whose Default?
Whether or not we have the data for documentation, it certainly seems true that writers nowadays compose shorter sentences than their counterparts in times past. It is not obvious though that this transformation is the result of any cognitive shift that has taken place among readers and authors. Shorter sentences may be easier in general for readers to process, and some less-accomplished readers may not be able to decode sentences whose length and structure go beyond specific testable boundaries. But we cannot assert that shorter sentences are either more or less natural than larger ones. Nor are they obviously better, even though we may sometimes prefer them.
In Writing Well, Donald Hall (1985) claims that brief senten:es are easier for everyone to write, whereas only skilled writers are likely to succeed at longer ones. This may indeed account for the emphasis in textbooks and writing guides on reducing long sentences, and it may explain the issues rookie writers face with passive structures. But if novices do not grapple with hard forms, they will never master them. While newspaper sentences may be extremely short, it is not clear that sentences in other genres of writing should conform to the same standard. True, the notion of readability has directed textbook". prose for twenty years or more. And the recent strength of the Plain English Movement has led to the simplification of many legal documents so that the average person may better comprehend them. But this is only part of the story, for despite Flesch's claim to the universality of his formulas, both the degree of readability and the plainness of language depend to a great extent on the content and presumed audience of a text.
The need to pitch college textbooks at a ninth grade reading level, apparently the limit of the average college student, has caused concern among writers and educators, who would prefer to raise reading levels, not simplify materials. And the translation into plain English of the technical language of contracts, leases, guarantees, and other official documents, which is part of a larger move toward consumer protection, represents a limited attempt to make these important aspects of modern life accessible to all the citizenry. No one seriously proposes to reduce every text to the least level of difficulty. Despite the passion to cut fat from sentences, the more sensible writing guides stress comprehensive revision rather than cutting alone, for good editing involves amplification as well as reduction. Even the computer style checkers, whose major concern is with the readability of technical documents, recognize the need to vary sentence length, and both Writer's Workbench and style IBM's program, Critique, allow users to define their own criteria for sentence length rather than accepting the default.
The movement of natural style in the direction of shorter words and phrases thus proves to some extent illusory. For one thing, we cannot with any confidence assert that modern writing is in fact shorter, more concise, or more natural in every context than writing in any ether age. For another, we recognize that a natural style does not come naturally. As Sir Philip Sidney knew four centuries ago, natural style is the art of making the difficult look easy. Claire Cook, in The MLA's Line by Line. How to Edit Your Own Writing, illustrates this by connecting passions today's for physical and syntactic fitness: "Trim sentences, like trim bodies, usually require far more effort than flabby ones."
It is what readers perceive as effortless, not the effortlessness of production, that brands a style as natural. Because readers bring to a text varying degrees of ability and experience, what is comfortable prose for one may prove difficult for another. Readability formulas that aim at the least common denominator will not solve this problem, for prose that is too easy can be as unnatural as prose that is too hard. One difficulty in constructing reading texts for use in schools is the fact that the simpler the words and sentences, the less interest the material has for students. A similar problem for writing instruction is the fact that naturalness is also a function of the appropriateness of language to the context of writing, and here no one can safely claim that less is always better than more, or vice versa.
