In medieval England, it was easy to tell the nobility from the peasantry: the nobility spoke French; the peasantry spoke English. This is why our words for meat on the hoof are Anglo-Saxon in origin (cow, sheep, pig) while our words for meat on the plate are French in origin (beef, mutton, pork). A peasant might steal a noble’s horse and cloak, but he could not pass himself off as a nobleman because he could not steal a noble’s tongue.
After the English crown lost its grip on its French possessions, and France became an enemy, the English nobility began to speak English. But they still needed a shibboleth to separate themselves from the peasantry, and so they created their own English, the English of the privileged and the learned. This English had a grammar based on Latin, which is why it forbade structures natural to English, but foreign to Latin, such as the split infinitive (to boldly go) or the ending of a sentence with a preposition (the sort of arrant pedantry up with which Churchill reputedly refused to put).
These particular latinate shibboleths are rarely insisted upon today in “standard” English, but grammatical shibboleths still abound; that is, linguistic conventions and usages that do not exist to convey meaning, but are recommended simply because they are “correct”. “Correct,” that is, not because they are necessary for communication to occur, but because they conform to an established standard.
Detecting which of our rules of grammar, orthography, and style are shibboleths is simple enough. If you can tell what the author of the offending passage meant, then any lapses from the standard are matters of shibboleth, not essential meaning. If you can “correct” the passage without any doubt as to is meaning, then you are correcting violations of shibboleths. The corrected passage does not mean something different from the uncorrected one, but it says (potentially) something different about the education, trustworthiness, or insider status of the author.
A resent WSJ article, This Embarrasses You and I*, focuses on the shibboleth function of grammar: the ways in which lapses from the standard in business communications may embarrass their authors. The joke here is that for the most part, these lapses from standard grammar do not actually embarrass those who commit them. Rather, they scandalize the sticklers who still see standard grammar as the mark of an educated person. These sticklers believe the lax should be embarrassed, but if the lax actually were embarrassed, they would change their ways. Since they don’t, we must presume they are not embarrassed. As the antics of the Hollywood train wreck of the moment teach us, the fact that we are scandalized by their behavior in no way implies that they are embarrassed by it.
In a Linked-in discussion (no link, because it is a discussion in a member’s only group) sparked by the WSJ article, Terry Kistner speaks for the sticklers:
Label me as an “Old School” stickler, but I believe the quality of grammer you use when communicating with others says a lot about your character. I suppose there is a time and place for slang or the use of acronyms, emoticons etc., and I can deal with that. I can even understand that there might be mistakes in a translation. But when someone uses “then” when it should be “than”, I want to pull out the rest of my hair. I think this kind of misuse says “I don’t care enough to know which word is the right word to use” or “I wasn’t paying attention in grammer classes”. Either way, I sense a lack of professionalism.
This is the essence of the stickler’s case: if you do not respect and obey the shibboleths of grammar, orthography, and style, people will not regard you as a professional. The function of these shibboleths is specifically to separate the professional from the unprofessional, just as classically their function was to separate the nobility from the peasantry.
The problem is that the ranks of professionals — people whose other credentials, accomplishments, and responsibilities mark them undeniably as professionals — no longer regard standard grammar as a hallmark of professionalism. (This much is clear from the WSJ article.)
To suggest that no member of the professional class will regard you as a fellow member of the professional class unless you follow all the shibboleths of grammar, orthography, and style, therefore, no longer holds water. These shibboleths are no longer the shibboleths of professionalism, they are the shibboleths only of sticklers (as Terry Kistner tacitly acknowleges by describing his view as “old school”).
The community of sticklers includes many writers (though clearly not all), and a scattering of readers who may be found in all sorts of different professional positions. I have encountered avid sticklers in the ranks of engineers and product managers. Often sticklers in these positions are sticklers even for some of the latinate rules that most contemporary usage gurus would regard as moribund. (In a sense, knowing that these latinate forms are now considered moribund is itself a kind of super-shibboleth, separating the grammatical elite from the common stickler.)
Anyway, my point is that, however much we might wish otherwise, adherence to standard grammar rules is no longer a shibboleth of either the educated or the the professional. People who demonstrate their professionalism and expertise in other ways are given a pass on lax grammar and awkward style. What are the consequences for technical communication?
First, it is worth acknowledging the undoubted truth that many of the linguistic conventions which would count as shibboleths by the test I proposed above, nevertheless play an important role in making communication more efficient. The familiar is easier to parse, so following familiar usage certainly helps lower the barriers to comprehension. This does not cover all shibboleth rules, since many of them are more honored in the breach than the observance, but it is important to acknowledge the role that conventional usage plays in easing comprehension.
That said, we have to acknowledge that one of the main criticisms that writers make of community and collaborative content — that is lacks authority because it violates grammatical shibboleths — does not hold water. The simple fact is that people do not judge the reliability of community content by its grammatical purity, a shibboleth that is meaningless to them. They judge it by other measures, such as the grasp of the subject matter that the author displays, or the respect that the author has earned from other members of the community. To believe that tech writers can or should ignore community content, or to believe that their careers will not be affected by its impact in the information space for a product, is to be blind to the reality of how trust is earned and granted on the Web.
We have to see and acknowledge that the community of technical information providers is now far broader than ourselves, and that quality in that community is not, except by a few, judged based on the shibboleths of language, nor by the more specific conventions and shibboleths of technical communication. And we have to understand and acknowledge that we will not enhance our own standing in the community by carping at the grammatical foibles of others. It will not make us look trustworthy; it will make us look petty.
Does this mean that we can or should abandon the shibboleths of language ourselves?
There is a principle in network information systems design: be liberal in what you accept and strict in what you send. In other words, a good network application will conform strictly to all network protocols in what it transmits, but will be as open as possible in what protocols and errors it accepts in the information coming from other systems. An example of the latter is a web browser, which accepts content from millions of web servers around the world and accepts, and makes the best of, all kinds of poorly-formed and error-filled HTML documents.
On the same principle, it seems like a good rule for us to be liberal in accepting community and collaborative content, while continuing to adhere to standard grammar and usage in what we ourselves produce (though not to the extent of being outright obscure in adherence to the obscure shibboleths still beloved of some sticklers).
Be liberal in what you accept, and strict in what you send. In a technical communication environment in which professional tech writers are only one voice in the choir, that seems like a pretty good operating principle to me. What do you think?