The Abstract and the Argument
How a scholar measures the truth in every qualifying verb
This paper examines how the linguistic resources of academic prose, far from being neutral instruments, encode a writer's calibrated relationship to the truth of a claim.
Drawing on a corpus of three hundred research articles, it argues that and boosting are not stylistic ornaments but the very grammar by which scholars distribute their intellectual liability.
Previous accounts have tended to treat as a matter of politeness, a softening of assertion designed to disagreement.
Such readings, this study contends, mistake a moral architecture for a social .
When a researcher writes that a finding may suggest rather than demonstrates a conclusion, she is not merely being courteous; she is measuring, to a fraction, how much of the truth she is prepared to be answerable for.
The central hypothesis is therefore straightforward, though its implications are not.
Register, understood as the systematic variation of language across situations, functions as an instrument of accountability rather than as decoration.
To test this claim, the analysis proceeds in three movements.
First, it establishes a of devices, distinguishing the modal, the approximative, and the frankly evasive.
Second, it correlates the density of these devices with the risk of the surrounding claim, as rated by expert readers.
Third, it considers the cases, those in which confident assertion and elaborate coexist within a single sentence.
The findings, presented in Section 4, are more troubling than the hypothesis anticipated.
Writers hedge most heavily not where the evidence is weakest, but where the professional cost of error is highest.
In other words, the grammar of caution tracks reputation more faithfully than it tracks doubt.
This is not, I hasten to add, an accusation of bad faith.
It is, rather, an observation about the quiet economy of self-protection that every disciplined prose style conceals.
The introduction that follows situates this argument within three overlapping literatures.
The first concerns the philosophy of assertion, and in particular the norms that govern when one is entitled to say that one knows.
The second draws on the sociology of science, which has long insisted that the laboratory and the journal are theatres of persuasion as much as of discovery.
The third, and the least expected here, is the tradition of literary stylistics, which has taught us to read a sentence's hesitations as we read a face.
It is at the of these three currents that the present study takes its stand.
A word, finally, about method and its limits.
No corpus, however large, can recover the private that produced a given hedge.
We infer intention from pattern, and pattern, as every reader knows, is a hypothesis and not a confession.
The claims advanced here are accordingly offered in the spirit they describe: with confidence to evidence, and with the boundaries of that confidence drawn in plain sight.
If the argument persuades, it will be because its own register has practised what it preaches.
If it fails, the failure will itself be , a tremor in the prose where certainty outran proof.
For the ethics of scholarship, I shall suggest, begins not in the conclusion but in the choice of every verb.
Every sentence, in the end, is a decision about how much of the truth one is prepared to be answerable for.
The chapters that follow are an attempt to read that decision, sentence by sentence, in the work of others and, unavoidably, in my own.