As everyone who has written them for a living knows, statutes are dense and complicated. Therefore, it is not surprising that examining a brief statutory passage will reveal a good deal about the art of writing statutes. To test this hypothesis let us imagine a drafter at work on a bill that creates a committee. It is necessary to say something about procedures, specifically about the need for a quorum. The drafter's initial plan is to write “there shall be a quorum.” However, that might not be the best course of action.

The drafter's initial reaction to that solution is uneasiness, which is exacerbated if he or she remembers that the primary function of statutes is to direct behavior. In essence, this first attempt at writing a statute is a direction to a quorum that it should exist. The drafter's unease is well founded. Neither the recipient of the direction nor the direction itself makes any sense. A quorum is an abstraction, not a human actor that can be expected to respond to a direction. Moreover, entities either exist or they do not exist; they cannot reasonably be expected to follow a direction to exist. The best clue to the second problem is that the verb is a form of “to be” not an action verb. That statutory sentence must be rewritten.

The best correction of that misstep is to create a conventional statutory condition and to connect it firmly to a behavioral directive. A common signal that a conventional condition is being created is the presence of "if," so let us begin a revision with that word. One possibility that appears to work is "if a quorum is present, the committee may conduct business." That is a significant improvement over the first attempt, but the drafter may still feel uneasy. The cause of that feeling is not immediately evident. However, the drafter ought eventually to realize that the situation in which a sentence about quorums is most likely to apply is one in which there is no quorum. That is the state of affairs when a question about whether the conduct of business by the committee is valid will occur. Recognizing that fact leads to the conclusion that an alternative that is still better is "the committee shall not conduct business if a quorum is not present." The drafter has moved from creating a condition precedent to creating a conventional condition and authorizing and then to creating a conventional condition and forbidding.

One indication that the drafter is on the right track is that he or she has shifted the decision maker in disputes about the validity of committee business as it relates to quorums. Under the condition precedent the decision maker is a court, which will do its well-known analysis of whether a statute is mandatory or directive. If the statute is mandatory, there are serious consequences if the statute is not followed: in this instance an action by the committee is invalid. If the statute is directory the only consequence of failing to follow it is that a legislative aspiration has been neglected: the action is still valid. In other words, a result-oriented judge has a free rein. Under the other two versions of the statute, the statute has determined that the lack of a quorum invalidates committee business. In other words, the Legislature is the decision maker. That is a result that is much to be desired. After all, drafters work for legislatures not courts.

One reason why the process that created this revision worked is that it did not operate in a vacuum. It is always tempting for a drafter to act as though he or she were working entirely in a structure of words. That is particularly true if the drafter is plagued by one of the profession's major problems, time pressure. To combat that outcome one must make frequent reality checks, noting the interaction of one's words with the real world. This progress has been described by Jacques Barzun in his essay "Lincoln the Literary Artist":

"Legal thought encourages precision through the imagining and the denial of alternatives. The language of the law foresees doubt, ambiguity, confusion, stupid or fraudulent error, and one by one it excludes them." One instance of those processes occurs in this essay at the shift to the final version of the sentence. That shift was a response to the realization that the important disputes over the sentence's meaning will occur when there is no quorum.

Another pattern that appears in the sequence of revisions of the sentence is increasing specificity. The connection of the quorum requirement to the validity of the committee's actions illustrates that development. One can imagine a spectrum that applies to various kinds of writing. At one end is the most specific writing, at the other end the most general writing. At the specific end judges have only a little leeway. At the general end they have considerable leeway, and the Legislature virtually loses control over the interpretation of the statute. As I have mentioned, drafters work for legislatures, so one of their goals should be to move their writing toward the specific end of the spectrum. That does not mean that all statutes must be exhaustive and that all interpretations should be literal. The latter is an impossible and unfortunate goal. There is room for judicial interpretations based on a statute's purpose when there has been a drafting error and when it was impossible for a drafter to imagine particular circumstances that might interact with a statute.

As the reality checks and the increasing specificity proceed through the sequence of revisions, material about the consequences of failing to have a quorum is added to the sentence. Some may argue that the draft's requester did not ask for this material and that it therefore is illicit. However, the drafter has acted within his or her ambit. The requester, if he or she even had a notion about quorums--which is far from certain—probably asked only in general for a requirement of a quorum, leaving it to the drafter to work out the details.  As all drafters know, requesters do not have specific intentions about all the details of a request.  In other words, the new material merely makes explicit that which was implicit in the first version of the draft.  It does not illicitly add unrequested material.

As the changes described above proceed, one thing that becomes very clear about the tactics described here is that they stimulate thought. This development will increase the accuracy of the draft, and accuracy, not clarity, should be the primary goal of statutory drafting.  One way to ensure that these benefits accrue is to choose the correct perspective from which to view the drafting process.  Three possible perspectives exist. One can look at the process from the perspective of the reader, the writer or the text.  Adopting the writer's perspective—asking what effect various tactics have on the writer—is the most efficacious choice.  That is, one should determine whether a tactic will help the drafter think effectively, not whether a tactic makes the reader's task easier.

Clarity, ease of reading, is the proclaimed goal of the Plain Language School. At this time it is useful to examine, in light of the test case, that school's effectiveness. An advocate of Plain Language would almost certainly react to the first version of the sentence by, without further thought, changing "shall" to must" and then moving on to the next problem, believing that the sentence has been made clear and nothing else need be done. That is, such a practitioner would run through his or her short list of rules for writing, find that the one about using words and phrases as they are used in ordinary speech has been violated because, he or she thinks, an ordinary person would use "must," not "shall." That is neither true nor relevant. My experience is that persons who want to impose a requirement make it either harsh or soft. That is, rather than using either "shall" or "must" they would say something harsh like "CLEAN YOUR ROOM" or something soft like "Can you get that report to me by noon?" More importantly, ordinary speech is irrelevant to drafting. That is because, as the philosopher Ludwig Wittgenstein points out, there are a myriad of language games, not one. Elizabethan sonnets are not written the same way as are telephone directories. Each language game has its own features, goals and appropriate tactics. For example, the statutory language game is noted for its strong dependence on conventions, its five functions, its goal of accuracy and other features.
In short, working with this test case has revealed a considerable amount about drafting. That kind of writing is enormously difficult and requires a considerable amount of sophisticated help. There is a growing body of material about drafting, but much of it is insufficiently sophisticated, partly due to the effects of Plain Language. Drafters need more analyses of their art that recognize its complexity.