The Pragmatics of Politeness in Scientific Articles

GREG MYERS
Uniwniry ofBrodford

Recent studies of the pragmatics of politeness have drawn on conversational

data. I argue that their model can be extended to some genres of written texts.
There have been two obstacles to such an extension: the lack of a d@nite
addressee for published texts, and the dificulty of d@ning relevant cultural
variables. Taking a corpus of articles by molecular geneticists, I assume a
simple model of a two-part audience, and focus on two kinds of impositions:
claims and denials of claims. With this framework, one can see politeness
strategies in regularities of scientific style-such as the use of pronouns and of
passives-that are usually explained in terms of conventions. The analysis also
accounts for some otherwise unexplained stylistic features, such as the use of
adverbs in establishing solidarity, and the use of personal attribution in
hedging. With these positive and negative politeness strategies in mind, we can
understand better the social significance of the occasional instances in which
the writer makes an imposition without redress, or makes the imposition
indirectly or chooses not to make it at all. Comparisons with popularizations, a
genre in which the writer has a different kind of relation to the reader, and thus
uses different kinds of politeness devices, show that these devices arise in
response to the interaction embodied in the text.

~~ ~~ - - - -

In the introduction to the new edition of their classic essay on the pragmatics of
politeness, Penelope Brown and Stephen Levinson (1987) survey some of the
many studies on the topic over the last ten years. I was surprised to find that
none of these studies deals with written texts. No one who studies either
conversation or written texts would want to minimize the differences between
the two, but I find that the model proposed by Brown and Levinson is useful in
trying to explain how we interpret some constructions found in writing, particu-
larly academic writing. Take, for example, one of Brown and Levinson's own
sentences.
However, where R-factorsand the influences of audiences are understood, and where
generalizations are taken from sufficient instances of observed language usage, it is
indeed possible to use the distribution of polite forms as a highly sensitive index of the
distribution of social equality and inequality, intimacy and distance, as shown (we
believe) for example by Brom (1 979, 1980), Levinson (1 982). (Brown and Levinson
(1987:12))
How are we to interpret the parenthetical remark, we believe? One assumes
they believe every statement in the book, unless they say otherwise or mark it as
AppliedLinpisrics, Vol. 10, No. 1 0 Oxford University Press, 1989.
2 THE PRAGMATICS OF POLITENESS

an example. The phrase does not seem to be used for emphasis, and it is not
serving the same function as the indeed, which is signalling that they are shifting
from criticisms of this assertion to a reaffirmation of it. I will argue that such a
phrase is inserted to mitigate what Brown and Levinson call a Face Threatening
Act (FTA). In their culture, the culture of linguists and anthropologists, it is an
imposition on other members of the field to cite one's own work without also
citing the work of others. For example serves the same function, by suggesting
that the work of others also supports this point, and their own articles just
happen to be handy. Such FTAs are inevitable in a text like their survey, because
it is a kind of interaction with other researchers in pragmatics and also with
those in the more general audience who are potentially interested in politeness,
and it is a kind of interaction requiring some impositions of the sort quoted here.
Their introduction, accordingly, has many examples of politeness devices.
But whose face is being threatened here? It is often hard, for a published text,
to say who is interacting and what interaction is involved. Anyone with £8.95
can pick up this text, and it is not addressed to anyone in particular, or even
restricted to linguists and anthropologists in general. If there is no immediate
speaker (S) or hearer (H), it is hard to define the variables used by Brown and
Levinson. Indeed, it may not seem to be an example of interaction at all, since
the people involved are not present to each other. But I think we can still use
some insights from conversational pragmatics as a guide to analysis of written
texts.1 If one has some idea of the social context of text like this, one can define
the possible audience fairly closely, and one can see requests, blamings, thanks,
and invitations in the text just as one can see these acts in the conversations in a
South Indian village.
Brown and Levinson present their study as part of the linguistic project of
showing universals in language usage; the striking parallels in politeness devices
between three unrelated languages shows that while the expression of politeness
may vary enormously from one culture to another, the basic hierarchy of polite-
ness strategies is not culture specific. Drawing on Erving Goffman's concept of
face (1967), they construct a system in which a Model Person is endowed with
Negative Face and Positive Face: roughly the want to be unimpeded and the
want to be approved of in certain respects (1987:58). The Model Person also
has a rational faculty for choosing the course of action that will give the highest
pay-off with the least loss of face, evaluating three variables: the social distance
between the Speaker and Hearer (D); the relative difference in power between
the Speaker and the Hearer (P), and the rank of the imposition (R). Brown and
Levinson argue that this basic system holds between widely separated cultures,
and that much of the variation between cultures can be accounted for in terms of
the variables they propose.
I will show some further examples of the strategies Brown and Levinson
describe, not from ethnographic notes of interactions but from published
writing. By finding the same sort of strategies in yet another subculture, I am
adding some evidence to their claim for the universality of these strategies. But
my project is not concerned with universals; my purpose is to use their model to
 GREG MYERS 3

help understand the interactions between writers and readers in written texts,
and particularly in scientific texts. I will illustrate politeness devices in writing
from texts that are not usually seen as involving interactions at all—a collection
of about sixty articles on molecular biology. Scientific articles are sometimes
treated as purely informational and impersonal, as collections of conventions—
the use of passives, nominalizations, hedges, acknowledgements—that can be
explained in terms of the norms of scientific culture. Following Brown and
Levinson's criticism of an analysis based on norms, I will argue that these
features can be better understood as rational strategies for dealing with the
social interactions involved in publishing an article. The same sort of analysis
could be done with business letters, agony aunt columns, book reviews, or the
last few years of Applied Linguistics. But these more obviously interactive text
types would make my argument too easy; I see scientific writing as the hard case,
and therefore the best place to start.
The main difficulty in analysing published writing in terms of politeness is that
there isn't just one Hearer, there is potentially a large and diverse audience.2
One way of dealing with this difficulty is by looking, not for the actual audience
of, say, Cell or Nature, and what their response might be, but by looking at
features of the text that suggest what its audience and its writer are intended to
be, how they are supposed to relate. Following this approach, we find not one
audience but two: the wider scientific community, to whom a research report is
supposed to be addressed, and an immediate audience of individual researchers
and particular groups of researchers doing similar work, who in a sense
'overhear'. That is, whoever really reads a scientific article, the forms are
consistent with it being addressed to a general scientific audience knowledge-
able in the research area. So, for instance, knowledge of some terms is assumed,
but well-known researchers and standard studies are cited as if the reader did
not know them. The writer may be thinking of just a few other researchers
reading the article, and he or she may not in fact get many more readers than
that, but the writer still addresses the general audience, and these readers, as it
were, 'listen in'. In Ludwik Fleck's terms, we can talk about an exoteric audience
involved in the ongoing research problem and an esoteric audience that takes an
interest in some of the researchers' findings. The distinction is important,
because politeness involves displaying to the exoteric group the proper respect
for the face of members of the esoteric group. And the same act may be an
imposition on one of these audiences but not on the other.3
In slicing the Gordian knot of audience in this way, I am making a strong
claim, not just that these two audiences must be considered, but that, for the
purpose of this analysis, we don't need a more refined or elaborate distinction.
There are of course many other possible audiences for a scientific text: science
journalists, administrators, biographers, sociologists of science. And there is a
whole field of literary study that looks beyond the individual reader to
communities and traditions of interpretation. But the only audiences that matter
here are those needed to account for features of the text. So while it is true that
some sentences from these articles are about to be read by linguists, it is not
4 THE PRAGMATICS OF POLITENESS

necessary to include these linguists as one of the potential audiences, because I

cannot find any features that can be explained by taking this audience into
account.
For scientific articles we also need to consider two different authors—the
voice we take as speaking in the text, and the researcher whose work is
described. We take this for granted when Brown and Levinson cite themselves,
in the third person, just as they would cite anyone else. As with audiences, the
reality is much more complex than this; all the research articles in my corpus
have multiple authors, and in each case there would be complex processes of
comment by colleagues, review by referees and editors, and embedding of the
writing of earlier articles. Again, there has been a great deal of interest in literary
and film criticism in looking beyond the individual author to the question of
authorship. And as we will see, writers of these scientific texts often present
themselves as merely speaking for their materials and instruments (This result
suggests...). But for our purposes, we can say that the text presents the author
as both writer and researcher. So we replace this dyadic relation:
Speaker Hearer
with this messier relation:
Author as Writer Readers
Author as Other
Researcher Researchers
When we look at other genres, besides the experimental report, we can see other
configurations of writers and readers.
To analyze the interactions of writers and readers, we need some model of the
cultural factors involved in scientific discourse that would explain, for instance,
why self-citation could be an FTA.41 will assume in the subculture of science: (1)
that the social distance between individuals—D—must be treated as very great;
(2) that the relative differences in power between individuals—P—are supposed
to be small, but (3) that the community as a whole is supposed to be vastly more
powerful than any individual in it. Thus we will often see the relations between
one researcher and another requiring little deference, while one researcher
must always humble himself or herself before the community as a whole. Of
course in reality scientists have a network of informal contacts, collaborations,
and long-standing personal commitments that do not require great social
distance. But none of this is to emerge in print: there are nofirstnames, and one
can even refer to oneself in the third person. And of course there are great
differences in relative power between a Nobel Prize winner and an assistant
professor at the University of Idaho, as observed at a conference, or in letters, or
in referees' reports. But in published discourse these differences in power must
never be acknowledged; everyone must present themselves as equally the
humble servants of the discipline. If we want to see how these variables, D and P,
vary, we would have to look at other genres, such as popularizations and news
reports.
 GREG MYERS 5
Scientific discourse consists of interactions among scientists in which the
maintenance of face is crucial. We can see scientists as building alliances that
define what knowledge is: the statement of the individual becomes a fact when it
is accepted and used by a consensus of the community (Latour 1987). In these
interactions, certain FTAs are unavoidable, and must be redressed with various
politeness devices. Every scientific report states a claim: in other words, it
makes a statement that is to be taken as the article's contribution to knowledge.
This is the statement that is implied when one cites the article. Most reports, in
stating a claim, deny or supersede the claims of others.-Consider the opening of
the famous 1953 Nature article by James Watson and Francis Crick that
announces the structure of DNA.
We wish to suggest a structure for the salt of deoxyribose nucleic acid (D.N.A.). This
structure has novel features which are of considerable biological interest.
A structure for nucleic acid has already been proposed by Pauling and Correy.' They
kindly made their manuscript available to us in advance of publication. Their model
consists of three intertwined chains, with the phosphates near the fiber axis, and the
bases on the outside. In our opinion, this structure is unsatisfactory for two reasons:
(...) (Watson and Crick 1953)

The claim here is that they have deduced the structure that should be accepted
by others in the field. The making of a claim threatens the general scientific
audience (the exoteric community) because it is a demand by individuals for
communally granted credit—in this case it is a bid for the Nobel Prize. The claim
also threatens the negative face of other researchers—the esoteric scientific
community—because it implies a restriction on what they can do now. The
sociologist W. O. Hagstrom treats such an offer of knowledge as a gift (in the
sense of Mauss) which is exchanged for credit. But this analysis underestimates
the importance of relations between competing researchers. One may recall
from Watson's best-selling memoir The Double Helix that Watson and Crick
did not read the articles of rival researchers as gifts; they dreaded that each new
issue of Nature or the Proceedings of the National Academy ofSciences (PNAS)
might contain an article by another researcher that would render all their work
useless. In the same way, if this structure they propose is accepted, no one else
can discover it, and any research based on other possible structures (if there
were any) would lose some of its validity. The making of claims always involves a
tension: the writer must stay within a certain consensus to have anything to say
to members of his or her discipline, but must also have a new claim to make to
justify publication. The weight of such FTAs can be related to the size of the
claim and the number of researchers who would have to alter their practices if it
were accepted. Here, Watson and Crick are making a weightier FTA than that in
Pauling and Corey's earlier article on the structure of DNA, because Watson
and Crick are proposing that not only biochemists, but geneticists as well, will
have to take their structure into account. The weight of the FTA of the claim is to
be measured in terms of social interactions, not in terms of any inherent size of
discoveries.
6 THE PRAGMATICS OF POLITENESS

There are other necessary acts that also threaten negative face—for instance
naming something, speculating, or asserting one's priority. To offer a new term
for a phenomenon discovered or created by the discipline is to invite the
question 'Who are you to name this stuff?' Speculation involves the assertion
that, although one does not have new data to offer the community, one's
thoughts are worth page space, general attention, and credit. It may also be seen
as an attempt to get an unfair jump on rivals. Priority disputes not only threaten
other researchers whose credit may be questioned, they threaten a community
that presents itself as being dedicated to the communal pursuit of knowledge
rather than to the individual pursuit of rewards. All these acts assert the
privileges of the individual and suspend the absolute authority of the group.
As soon as Watson and Crick assert their claim, they go on to dismiss the
structure proposed by Pauling and Corey as 'unsatisfactory'. Such dismissals
obviously threaten the positive face of the particular researchers being dis-
missed. One may recall from The Double Helix that in private, informal
conversations Watson gleefully mocked the Pauling and Corey article; here the
criticism is mitigated with the phrase 'In our opinion...'. (As we will see, in these
articles any attribution of a statement to a person weakens it.) Outright criticism
in print is so threatening that it is usually avoided, and rival claims are usually
briefly denied or even ignored. Since scientific claims survive only by being used
and cited in later research, to ignore a claim is to kill it, and it is the omission of a
reference rather than outright attack that is the main threat to an individual's
positive face. The denial of an entrenched claim is likely to threaten, not only the
researcher who proposed it, but a broad range of readers who have included
the challenged claim as one of the assumptions of their research. The more
researchers using a claim, the greater the FTA in denying it. No one had done
anything with the Pauling and Corey model, so, in this case, the denial of the
model will threaten only its authors. But when articles were published several
years later proposing alternatives to the Watson-Crick structure, they
threatened not only Watson and Crick, but hundreds of other molecular
biologists already working on the genetic code suggested by base pairing.
My own claims are that the acts of claiming and denying claims will usually be
redressed (in Brown and Levinson's term) with some sort of device to make
them more polite, and that these politeness strategies indicate the impositions
inherent in these acts. That is, the texts themselves show that the authors are
aware of the FTAs. These devices are not always specific words or phrases that
signal politeness: one could not find them just by searching for certain tokens.
Rather, politeness is 'implicated by the semantic structure of the whole
utterance' (Brown and Levinson 1987:22). We can find in articles on molecular
biology many of the strategies listed by Brown and Levinson. I will focus on
some positive politeness strategies that emphasize the solidarity of writer and
reader, and on some negative politeness strategies involving various kinds
of hedging.5 hi both cases, the analysis of politeness leads us to some features
we expect of scientific writing, features that can be explained well enough by
means of other analytical frameworks. But an understanding of politeness also
 GREG MYERS 7

accounts for some features of scientific writing that are less often remarked
upon, but that occur frequently. Most of the politeness devices I have en-
countered in this corpus work as either positive or negative politeness, but there
are also some instances of the other choices Brown and Levinson consider in
their hierarchy of strategies: doing the FTA baldly, without redress; doing it off-
record, by implication, and not doing the FTA at all. The situations in which
authors choose these less-preferred strategies show the kinds of reasoning they
employ in choosing the phrasing of claims and denials.6

POSITIVE POLITENESS AND SOLIDARITY

In Brown and Levinson's hierarchy, positive politeness strategies are directed at
showing the speaker's concern with the hearer's concerns. In scientific writing,
the range of possible references to the readers' wants is severely restricted: one
cannot, for instance, make any remark praising the general talents of a
researcher, or remind the readers of a researcher's past successes. But there are
positive politeness devices for showing the writer's acceptance of the wants of
rival researchers, or of the scientific community as a whole. We would expect to
find, and we do find, strategic use of pronouns to stress solidarity as an
imposition is made (Brown and Gilman 1960). But we also find some features
we might not notice except in the analysis of politeness: the use of modifiers
assuming common ground; the use of emotional responses to indicate solidarity,
and such unscientific-seeming devices as joking and giving gifts.
Positive politeness devices are used to mitigate both claims and denials of
claims. One way of making a criticism while minimizing the FTA is for the
writers to use pronouns that include themselves in the criticism. Besides the we
that means the writers, there is a we that means the discipline as a whole. When
W. F. Doolittle criticizes an evolutionary misconception held by many in the
field, he uses we rather than they.
Although we may thus view genes and DNA as essentially "selfish", most of us are,
nevertheless, wedded to what we will call here the "phenotype paradigm." (Doolittle
and Sapienza 1980)
Here the second we means Doolittle and his co-author Sapienza, while the first
we and the most of us means everyone in biology. Crick uses the same device
when drawing lessons from the split genes episode.
Lacking evidence we had become overconfident in the generality of some of our basic
ideas. (Crick 1979)
Such criticisms are only allowable when put in this general form; a sentence like
'Lacking evidence, John Smith became overconfident...' would be very rude
and very unlikely. The pradox, of course, is that the writers are identifying
themselves with the very ideas they are refuting—usually, as here, by showing
how the field has changed. It is not unusual for a writer to list a number of
unsatisfactory views before going on to a new synthesis. Here Blake has listed
his own earlier articles among others before dismissing them all.
8 THE PRAGMATICS OF POLITENESS

Thus none of the current ideas on the relation of coding sequences to protein function
and structure seems fully correct. Rather than simply discarding them, however, it may
be possible to place them in a more comprehensive context which takes into account a
range of other studies. (Blake 1983)
Note that in this case it is not a question of the use of pronoun to suggest
solidarity, but of a pattern of citations (he has cited his own earlier articles along
with others) that has a similar effect.
Writers can show their solidarity with the community more subtly by
exhibiting responses that assume shared knowledge and desires. All the
instances I can find of an indication of emotional response to results, or desire
for certain results, show identification with a common goal, rather than the
responses or desires of an individual. In fact, there seem to be only a few
emotions in this body of texts: surprise at a new discovery; satisfaction at the
progress of the discipline, and disappointment at the failure of a received idea.
For instance, researchers who made the initial discovery of split genes express
their amazement in the title of their first article,and the word keeps recurring in
later accounts.
'An Amazing Sequence Arrangement at the 5 ' Ends of Adenovirus 2 Messenger
RNA'(Chowe/a/. 1977)
The word amazing is not an assertion of the author's own personal feelings,
which would be a grave FTA, but an expression that they were as surprised to
make this discovery as anyone else would have been. In other words, they are
not claiming any special foresight for themselves. Similar expressions mark the
claims of a number of other articles.
Unexpectedly, we have found that the DNA sequences complementary to ovalbumin
mRNA are split... (Breathnach et al. 1977)
When a researcher expresses pleasure, it cannot be at the triumph of his own
group, but only at the way a finding fits in with the larger goals of other
researchers who might be reading the article.
One pleasing feature of this structure is that the 3 ' terminal G-OH is not locked in an
interaction that would preclude it from reacting with the cyclization site at A16 (3).
(Cech etal. 1983)
In the same way, unhappy emotions are used to express what are assumed to
be the feelings of most researchers at a promising hypothesis or assumption
crumbling under experimental evidence. Crick uses this device frequently in
summarizing the state of research.
Such a hypothesis would predict that split genes would not be found in mirochondria.
Unfortunately, the experimental evidence suggests that there, too, genes are split into
pieces. (Crick 1979)
Doolittle uses the indication of emotion to suggest an identification with the
established paradigm—that is, with the very position he is criticizing in his
article.
 GREG MYERS 9

The evidence for a phenotypically functional role for middle-repetitive sequences

remains dishearteningly weak 40 " 43 ... (Doolittle and Sapienza 1980)
Here the identification with the view criticized is a politeness device showing
that the writer wants what most readers want. The same sort of device can be
used to redress direct criticism in the discourse of other disciplines in which
direct criticism is more common. For instance, when Brown and Levinson
criticize Robin Lakoff, they phrase their sentence in a way that stresses their
desire to agree with her.
Indeed R. Lakoff has suggested (1973a) that this is the most important feature of
politeness. In this we cannot agree with her, yet it is true that in certain circumstances-
bothering important persons for favours, for instance—this is an important element.
(Brown and Levinson 1987:130)
As soon as they have registered disagreement, Brown and Levinson go on to
point out where they do share common ground. (In this case, as it happens, the
authors are dealing with a researcher to whom they say they owe a personal and
scholarly debt, but the same phrasing could be used to disagree with anyone,
and indeed it would be an FTA to treat Lakoff with any special consideration.)
A speaker at the autumn 1987 meeting of the Linguistics Association of Great
Britain said that' Unfortunately, I don't think that many of the correlations that
are supposed to hold with null subjects actually hold water'. The speaker was
not, in fact, sympathetic with Chomsky's approach, nor was the audience, but he
chose to mitigate his criticism by assuming the point of view of those he was
criticizing.
One can also identify one's work with that of the whole field by assuming
everyone shares either the idea behind one's claim or the experience behind
what might be taken as criticism. For instance, Blake introduces a speculative
claim by saying:
In the light of these observations I wish to consider the consequences of the rather
obvious assumption that exonic regions of DNA ... correspond to integrally folded
protein units ... (Blake 1978)
By treating the assumption as 'rather obvious', Blake includes all his readers as
potentially capable of making this claim, minimizing his own originality. (Of
course, if the assumption and its consequences really were 'obvious', the article
would not be publishable.) A similar technique can be used in presenting
potentially face-threatening comments on the work of another researcher.
Chambon comments on the way Leder interpreted his results before he heard
other speakers at the Cold Spring Harbor symposium and changed his mind.
In retrospect, it is amusing to note again the way we interpret an experiment is strongly
influenced by what we already know. For instance, at the meeting, Leder et al.
interpreted their R-loop data as indicating the juxtaposition of two non-allelic beta-
globin genes. (Chambon 1978)
It is surely a threat to Leder's positive face to bring up an erroneous,
unpublished version of results he later published with a quite different (and very
10 THE PRAGMATICS OF POLITENESS

important) interpretation. Chambon mitigates the impoliteness by including

everyone in this behaviour of changing their minds (and indeed, he points out in
other articles how his own group reinterpreted their results in just the same
way). Within this group such a shared experience, which might otherwise be
seen as embarrassing, is 'amusing'.
Joking would seem to be an unlikely politeness strategy in scientific texts, and
indeed there is not much room for humour in PNAS or Cell. But jokes are
useful for scientists, as for linguists and other academic communities, in
establishing a sense of shared knowledge or assumptions, and this sense can
emerge in two features of scientific articles at least: new terms and titles. In
molecular biology, as in other scientific disciplines, highly speculative and thus
highly threatening ideas are often presented with half-serious coinages that
seem to disarm a threat. So, for instance, Orgel and Crick (1980) write about
the DNA that has no apparent function as 'junk' DNA, first in quotation marks
and then, without the marks, as a regular term. This is the same Crick who
earlier coined such terms as the wobble hypothesis and the Central Dogma,
half-joking expressions that have since gone into the textbooks. In other
disciplines there are such well-known coinages as charm, colour, and quarks in
particle physics; the selfish herd hypothesis in sociobiology, and fuzzy categories
in linguistics.
A joking title may serve to mitigate the FTA of a claim. The first announce-
ment of the remarkable and rather unsettling discovery that led to all these other
articles I studied is made under an unconventional title with a punning use of
architecture: 'Adenovirus-2 messengers—An example of baroque molecular
architecture'. One of the other articles in this particular group also has a punning
title: 'Making ends meet: a model for RNA splicing in fungal mitochondria'. It
may be that such jokes indicate something about the authors, but they may also
indicate something about the situation. Both these articles present far-reaching
claims at more or less the same time as other competing laboratories: the joke
suggests the shared amusement of the esoteric group rather than competition
for credit. A glance through the Cold Spring Harbor symposia show that such
titles are common, though not as common as the playfully self-referential titles
in linguistics. We find them in proceedings of conferences, not only because
such forums are less formal, but also because it is more important in this
situation to establish a link with the immediate audience that attends the
conference than to provide an indexable summary title that includes all the
necessary keywords.7
Brown and Levinson list the giving of gifts as one of their positive politeness
strategies. This device may seem to have more relevance to American Indians
than to molecular biologists. Molecular biologists do give each other DNA
samples and sequencing expertise and the like, and research could not continue
without these exchanges. But the kind of gift that concerns us in the analysis of
writing is the gift of credit, especially in the disposition of citations. It may seem
odd that the choice of who to cite and who not to cite could be a matter of
strategy, but there are always far more papers that could be cited than the
 GREG MYERS 11

journal would allow, and selections have to be made. 8 One special case of this
requirement of giving credit is the acknowledgement of simultaneous,
independent claims. Historically, priority disputes have been major causes of
tension within the scientific community. For instance, there has been an
enormous amount of discussion of the issues concerned with the discovery of
DNA. But Watson and Crick avoid any suggestion of rivalry in a brief (and now
famous) acknowledgement:
We have also been stimulated by a knowledge of the general nature of the unpublished
experimental results of Dr. M. H. F. Wilkins, Dr. R. E. Franklin, and their coworkers at
King's College, London. (Watson and Crick 1953)

In the texts I am considering, nearly every major claim was made by two or
three labs at about the same time, and communicated through the esoteric circle
of researchers long before publication. The claims were then sifted out, within a
year or so, so that one article could stand for all of them. It is important that
these first articles demonstrate that the writers have no intention of disputing
priority, and yet they do not want to say that their own claim follows and is
subsidiary to that of the other labs. So a formulaic sort of statement often
appears in the last paragraph.
We have recently learned of similar experiments by Berget, Moore, and Sharp (1977),
who used electron microscopy to examine hybrids between purified hexon mRNA and
single strands of DNA. They observed that the 5' terminal mRNA sequence appeared
as a single-stranded tail, which was complementary to three noncontiguous regions of
the Ad2 genome with map coordinates essentially identical to those reported here.
(Chow etal. 1977)
After this work was completed, we learned that Michel and Dujon (40) and Waring et
al. (41) have also proposed secondary structures for the Tetrahymena IVS and
adjacent exon sequences. Their structures, which are based on homology with yeast
mitochondrial introns rather than the use of structural probes, are similar but not
identical to ours. We find it encouraging that two quite different approaches have
converged on a common structure. (Cech etal. 1983)
While this manuscript was being prepared, a note appeared in Nature (London) [272,
581 (1978)) by W. F. Doolittle which proposes the same general premise as this
article—that splicing and "genes in pieces" is representative of an early state of cell
evolution. Doolittle then goes on to advocate the idea that... During the review of this
manuscript a second note [C. C. F. Blake Nature (London) 273,267 (1978)] appeared
welcoming the proposal that... (Darnell 1978)

The similarity in form among these many acknowledgements suggests the

interactional problems of dealing with priority. On one level, all of these
acknowledgements display politeness in crediting the rival researchers and
describing their work, whether it is already published or not. Never, in all the
many occurrences of this device, are there any instances (such as one might find
in other disciplines) of a writer using such an acknowledgement to point out
weaknesses in the rival approach. On the other hand, the opening phrase of such
12 THE PRAGMATICS OF POLITENESS

an acknowledgement is always a statement of how far one was with one's own
work when one heard about the other work, so the first statement, and the
background for the rest, is that one developed this claim independently. The last
part of such an acknowledgement is often a statement of satisfaction at how well
the results agree and support each other, reaffirming the communality of
scientific knowledge. So the acknowledgement is a highly formulaic device for
converting a possible conflict into the appearance, at least, of co-operative
effort.
I have called Brown and Levinson's study a 'classic essay'. Experimental
reports do not usually allow this sort of evaluative comment. When such
evaluations are made, they are more likely to be made when the author is
disagreeing with the researcher praised. So, for instance, Tonegawa refers to
Rabat's work as 'a leading model for the behavior of the hypervariable regions'
before going on to give experimental evidence refuting this model. As we will
see, authors of reviews, news articles, and popularizations have more leeway, so
they can praise the skills of fellow researchers in order to single them out for the
attention of those outside the inner circle.
Brown and Levinson say that the risk of positive politeness is that the Hearer
might not accept the Speaker as one of the set of people from whom positive
politeness of a certain form is welcome. Positive politeness is also risky in
scientific articles, but for a different reason: any indication of a personal interest
or a personal relationship conflicts with the duties owed to the community as a
whole. So personal praise, even as a politeness device, threatens both the one
praised—the one who is singled out for personal comment—and the one who
praises.

NEGATIVE POLITENESS AND HEDGING

Given the large social distance we have assumed between participants in
scientific articles, we would expect to find in them what Brown and Levinson
call negative politeness, the strategies assuring the readers that the writers do
not intend to infringe on their wants, their freedom to act. Most of the features
that are considered just conventional in scientific texts—hedging, impersonal
constructions, the assertion of general rules—can be reinterpreted as negative
politeness devices. But there are also some devices that are not usually noticed
as characteristic of scientific writing, for example constructions emphasizing
personal point of view.
Hedging is a politeness strategy when it marks a claim, or any other statement,
as being provisional, pending acceptance in the literature, acceptance by the
community—in other words, acceptance by the readers. So Watson and Crick,
in their claim, suggest a structure for DNA, and it is offered just as a structure,
not as the structure. In their next sentence 'considerable biological interest' is an
understatement, coming from a researcher who had celebrated their discovery
by telling everyone that they had found the secret of life. The hedging reflects,
not the probability of the claim, and not the personal doubt of Watson and Crick
(who were not famous for caution or modesty), but the appropriate attitude for
 GREG MYERS 13

offering a claim to the community. As soon as the claim is properly introduced,

the rest of the article can proceed without constant hedging (as Watson and
Crick's description does). And as soon as the claim is part of the literature, it
becomes possible to refer to it without any hedging.
The hedging of claims is so common that a sentence that looks like a claim but
has no hedging is probably not a statement of new knowledge. The abstract of
one article begins with what certainly sounds like a claim:
Splicing of theribosomalRNA precursor of Tetrahymena is an autocatalytic reaction,
requiring no other protein in vitro. (Cech etal. 1983)

But the author can make this statement baldly only because it was claimed in an
earlier article and accepted by other researchers. The claim of this same article,
the last sentence of the abstract, is mitigated by impersonalization and hedging:
Thesefindingssuggest a common origin of some nuclear and mitochondrial introns and
common elements in the mechanism of their splicing, (op. cit.)

We can tell this is the claim, not only by its position, but by the provisional
nature of the statement.
Hedging can be done with a modal verb making a conditional statement
(would or could) or with a modifier (probably) or with any device suggesting
alternatives (such as the indefinite article in the Watson and Crick claim)—
anything but a statement with a form of to be that such and such is the case. A
typical discussion section of an article (the conclusion that may draw out the
theoretical implications of the results) includes many of these devices. Here the
writers from Sharp's laboratory, having announced a startling discovery, are
discussing its implications, and setting forth a model to explain it.

Thus, a common short sequence of RNA might be attached to several mRN As. This is
consistent with the observation of R. Gelinas that...
The three short segments... are probably spliced to the body of this mRNA...
The region of the genome coding for the body of the hexon mRNA and the sequences
in these three short RNA segments in the 5' tail of this mRNA are probably included
in this long transcript.
Thus, a plausible model for the synthesis of the mature hexon mRNA would be...
This would result in the maturation of one mRNA species ...
• It is interesting to speculate on how general such a model for the processing of
eukaryotic mRN As could be.
Assuming that... this mechanism would certainly explain the observations ...
This RNA segment could be involved in the selection of certain viral RNA sequences
for transport to the cell cytoplasm or could be responsible for the preferential
translation of viral mRNA during the last stage of infection. (Berget, Moore, and
Sharp 1977)

In French, similar hedging can be done with the conditional.

Cette proteine serait douee de propriete d'autonomie intrinseque de son ARN
messager. (Jacq etal. 1980)
14 THE PRAGMATICS OF POLITENESS

The be-\erbs are generally avoided in making claims; a more common way of
phrasing them uses a verb like suggest:
These findings suggest a common origin of some nuclear and mitochondria! introns
and common elements in the mechanisms of their splicing. (Cech et al. 1983)
These results suggest that Ul RNP is essential for the splicing of mRNA precursors...
(Padgett et al. 1983)
All molecules of the snRNAs ... appear to exist in the form of antigenic snRNPs.
(Lerner etal. 1980)
(I will discuss the impersonal subjects a little later.) Another form of hedging is an
introductory phrase that tells the readers what weight the statement is to have.
It is not known whether this phenomenon also occurs in eukaryotic cell mRNA.
However, because of the limited coding capacity of the SV40 genome, it is likely that a
cellular process is responsible for generating the 5' spliced end of SV40 mRNA. (Hsu
and Ford 1977)
This kind of framing is used especially to suggest degrees of doubt about an
alternative interpretation of results or theoretical possibility.
Thus, it seems highly likely that RNA-RNA splicing is truly the mechanism for
bringing "mosaic" mRNAs together... (Darnell 1978)
Since it seems improbable to us that the sequence present at the 5 ' end of many of these
late Ad2 mRNAs is actually coded by the host genome and is only complementary to
these three Ad2 sequences by chance, we believe that these sequences are probably
transcribed... (Chow et al. 1977)
It is unlikely that the inhibition of splicing following preincubation with either anti-
(U 1)RNP or anti-Sm sera is a nonspecific effect due to formation of an antibody-RNP
complex. (Padgett et al. 1983)
It seems totally implausible that the number of radically different genes needed in a
salamander is 20 times that in a man. (Orgel and Crick 1980)
Each of these phrases, and dozens of others like them, mark the writer's
commitment to or rejection of a statement while formally leaving open the
(slight) possibility that the writers will be judged wrong.
It may seem paradoxical that in redressing some FTAs, there is some
reference to personal opinion or desire ('it seems improbable to us';' We wish to
suggest'), while in redressing others the personal subject is replaced by some
impersonal actor such as 'findings' or 'results'. It is of course true that scientific
writing often uses impersonal constructions and passives. But it is not true that
personal subjects are eliminated; the subject we (and, less often, / ) plays an
important role in politeness. To see how this works, we must recall that scientific
knowledge is supposed to be taken as universal; therefore any implication that a
belief is personal weakens it. This is not necessarily the case outside science; as a
response to an announcement of engagement, 'I think that's wonderful' is at least
as strong as "That's wonderful'. But consider, for instance, a sentence from
 GREG MYERS 15

Brown and Levinson's response to Leech. If they were just to repeat his
statement.they would imply agreement with it.
That the CP and PP are coordinate principles is shown... by the fact that without the PP,
the CP would make erroneous predictions.. .(Brown and Levinson 1987:4 [modified])
When they insert into this statement the personal attribution, 'he claims', they
signal their disagreement:9
That the CP and PP are coordinate principles is shown, he claims, by the fact that
without the PP, the CP would make erroneous predictions...
Brown and Levinson are doubting a statement by someone else. But the same
technique of giving a statement personal attribution can act as a hedge on one's
own claims and denials of claims.
We believe the functioning gene in the myeloma will consist of the precursor region
followed first by the 93-base-long interspersed DNA, followed by the variable region
gene, then by a 12 5 0-base-long piece of non-coding DNA (4), and last, by the constant
region gene. (Tonegawa et al. 1978)
/ would like to argue that the eukaryotic genome, at least in that aspect of its structure
manifested as "genes in pieces" is in fact the primitive original form... / would suggest
that DNA replication and transcription as well as translation would probably have
been unfaithful in the last common ancestor of prokaryotes and eukaryotic nuclei.
(Doolittle 1978)
We shall assume, for the rest of this article, that this hypothesis is true ...We can make
one prediction on the basis of energy costs. (Orgel 1980)
... nous avons trouve un resultat qui nousparait interessant... Bien des points restent
obscur dans l'expression des genes morceles mais l'idee d'une nouvelle classe de
proteines douees de propriete d'autotomie intrinseque de leur ARN messager merite, a
notre avis, d'etre envisagee d'une maniere plus generate en biologic (Jacq et al. 1980)
As with other hedges, the form of the statement reflects a relation between the
writer and the readers, not the degree of probability of the statement. For
instance, these three claims are all introduced with personal attributions, but
they vary from the almost certain to the wildly speculative.
We conclude that splicing activity is intrinsic to the structure of the RNA, and that
enzymes, small nuclear RNAs and the folding of the pre RNA into an RNP, are
unnecessary for these reactions. We propose that the TVS portion of the RNA has
several enzyme-like properties that enable it to break and reform phosphodiester
bonds. (Kruger etal. 1982)
From a close inspection of their conformations by drawing various stereodiagrams and
the cal<>lla-Cal|'ha distance map, / now propose a conformational characterization of the
segments as structural units . (Go 1981)
/ adopt the attitude that in most cases this is because viruses are short ofDNA and, by
various devices, their limited amount of DNA is made to code for more proteins than
would otherwise be possible...
16 THE PRAGMATICS OF POLITENESS

Should a chromosomal gene whose transcript was processed to make more than one
protein, / would expect that in the course of evolution the gene would be duplicated

/ have been so rash as to say. more than once, that we might expect between 10 and 100
different enzymes; but that was pure guesswork. The number could be as low as two.
(Crick 1979)
In the Kruger article, the personal statements have the effect of marking the
conclusions to be drawn from a complicated series of experiments. Go's
personal statement marks a theoretical model. Crick's apparently casual guess is
part of an article in which extensive speculation is mitigated by constant
demonstrations of the solidarity of the writer with the readers; we are included
as part of a group tossing around ideas in the wake of the discovery. So there is
no obvious correlation between the form of hedging and the degree of certainty
(though a phrase like 'that was pure guesswork' would seem to leave lots of room
for doubt).
Watson and Crick introduce their criticism of Pauling's model with the
phrase, 'In our opinion'. The same sort of hedge by personal attribution is often
used when criticizing the views of other biologists.
We think it has not been sufficiently emphasized that non-phenotypic selection may
inevitably giveriseto transposable elements and that no phenotypic rationale for their
origin and continued existence is thus required. (Doolittle and Sapienza 1980)
In the quotation with which I opened this paper, when Brown and Levinson say
parenthetically 'we believe', they are using a politeness device to mitigate the
FTA of their self-citation, saying that they believe that their later works support
their earlier argument, but that the readers are still allowed to judge for
themselves. Similarly, they can use this sort of hedge when criticizing other
interpretations, implying that they are willing to have the readers choose the
more persuasive explanation.
Robin Lakoff s analysis of this... seems to us counterintuitive. It seems more accurate
to see... (Brown and Levinson 290n)
A special case of this kind of hedging by personal attribution occurs with the
coining of terms. It is an FTA to try to appropriate some phenomena by giving it a
name oneself. We can see the dangers when other researchers resist a coinage;
for instance, some researchers still do not use the term intron, though it was
introduced ten years ago and is now used in the textbooks: they prefer the
alternative terminology of intervening sequences. So the introduction of a new
term has to be marked with some sign that the writer suggests it only provision-
ally, subject to its adoption by the community. Walter Gilbert introduced the
terms intron and exon twice, in articles in PNAS and Nature, using slightly
different forms of statement. In PNAS, it is as if these forms were used at his
laboratory and Tonegawa's as a matter of convenience.
We call such an additional piece of DNA that arises within the gene an intron (for
intragenic region or intercistron) and dius look upon the structure of this gene as ...
(Tonegawae/a/. 1978)
 GREG MYERS 17

In Nature, Gilbert makes an explicit bid to have these terms accepted generally,
as a convenience for everyone, replacing terms he considers inadequate.
The notion of a cistron . . . now must be replaced by that of a transcription unit
containing regions that will be lost from the mature messenger—which / suggest we call
introns (for intragenic regions)—alternating with regions which will be expressed—
exons. (Gilbert 1978)
Gilbert still hedges his proposal by saying he only suggests these terms. But he
suggests them more directly in this less formal and frankly speculative 'News
and views' article than in the PNAS report of experimental results. The same
sort of hedging of a new term by accepting personal attribution is used by Cech's
group, and Slonimski's.
Because the IVS RNA is not an enzyme but has some enzyme-like qualities, we call it a
ribozyme, an RNA molecule that has the intrinsic ability to break and form covalent
bonds. (Kruger etal. 1982)
Nous I'appelerons l'((ARN-maturase)). (Jacq etal. 1980)
Notre postulat est qu'il existerait deux classes de proteines que nous designerons par
ew-proteines et m -proteines. (Slonimski 1980)
This is consistent with the kind of phrasing used earlier to introduce what are
now key terms in molecular genetics.
This genetic unit of co-ordinate expression we shall call the "operon". (Jacob and
Monod 1961)
Literally, such formulations with personal subjects only say that they use these
terms, not that they would have others use them.
Brown and Levinson note that several kinds of impersonal constructions can
be a form of negative politeness, as when one says 'One shouldn't smoke in this
section' rather than 'Don't smoke here'. We expect such forms in scientific
writing: my argument here is that they do not necessarily reflect the im-
personality of science, or the arbitrary dictates of tradition, but rather are
rational ways of dealing with interactions, with claims and denials, the necessary
FTAs of scientific writing. So as we have seen, the most common form for a
statement of a claim attributes it to some impersonal agency:
These observations suggest that both mRNAs contain a long common sequence ...
(Klessig 1977)
These results imply a novel mechanism for biosynthesis of SV40 mRNA. (Aloni et al.
1977)
Thisfindingsuggests that the precursor mRNA is processed through the removal and
rejoining of internal RNA sequences. (Tilghman et al. 1978)
However, the sequences in these boundary regions share common features; this leads
to the proposal that... (Breatnnach et al. 1978)
If the assumptions of ultraviolet transcription mapping are correct, the data in this
paper imply that late during Ad2 infection of HeLa cells, there exists one or more
18 THE PRAGMATICS OF POLITENESS

closely spaced promoter(s) needed for the transcription of the right two thirds of the
genome. (Weber et al. 1977)
Ces consultations: 1) montrent que des fluctuations climatiques naturelles de la
temperature peuvent exercer une influence directe important sur les Reptiles meme
dans des regions temperees: 2) suggerent l'existence, chez les Viperes, de cycles
endogenes plusrigidesqu'on ne le pensait. (Saint Girons 1981)
The same strategy of impersonating can be used when challenging statements
by other writers; the authors can phrase doubts so as to avoid attributing the
criticized position or their own position to anyone. So, for instance, Darnell
devotes much of his paper to a contrast between two views of evolution, but he
does not attribute the view he is attacking, and he proposes his own view as an
impersonal alternative.
A contrasting view for the origin of divided genes ...
This view denies an intermediate role...
According to this view ...
In this scheme ... (Darnell 1978)
When Doolittle criticizes a received idea, he cites examples of papers relying on
the assumption, but he uses a string of passive sentences, avoiding attribution in
the text:
Untranslated messenger RNA sequences which precede, follow, or interrupt protein-
coding sequences are often assigned a phenotypic role in regulating messenger RNA
maturation, transport or translation. (Doolittle and Sapienza 1980)
Similarly, when Blake questions the applicability of Mikito Go's analysis of
haemoglobin, he does not say 'Mikito Go claims that . . . ' which would be
impolite in its direct suggestion of personal error, but uses a more elaborate
passive form, in which her name figures just as a technical term:
This is equivalent to a "Go plot" in which the "valleys" along which the splice junctions
are claimed to run correspond to minima in the distance sum. (Blake 1983)
The criticism is still there, but it is made indirectly. (This is one example where
there is no word or phrase that could by itself indicate politeness.) While
scientists must be careful to acknowledge all possible rivals, they must also try to
avoid attributing the views they attack to any one reasearcher.
Hedging, personal constructions, and impersonal constructions are the main
forms of negative politeness found in these articles: all of them work by
indicating the writer's deference before the scientific community. There are
several other devices that, while they look different linguistically, follow the
same general reasoning. For instance, Brown and Levinson note that one way of
making an FTA impersonal is to invoke a general rule: 'Bicycles may not be
carried on trains departing between 0700 and 1100'. When Doolittle criticises
one account of the evolution of split genes, he invokes one kind of general rule
that is likely to be accepted by all biologists, the main principle of Darwinian
theory:
 GREG MYERS 19

The idea that cells would do so to increase their potential for future evolution is not a
Darwinian one. (Doolittle 1978)

This is equivalent to saying the idea is wrong. Francis Crick invokes the same
general rule in a slightly different linguistic form:
But one should not evoke some selective advantage occurring only in the future ...
(Crick 1979)

Thus it is not Crick or Doolittle who is criticizing these proposals, but the
general and accepted rule. Another general rule frequently invoked is a
methodology requiring falsifiable hypotheses open to testing. Within the Poppe-
rian rhetoric of scientific method, the offer of speculation is personal and isolat-
ing, while the proposal of tests allows its integration into the disciplinary
structure of knowledge. This is the way speculation is offered in the wake of
some startling new experimental results:

It is too early to speculate on the role of the split gene organisation in the regulation of
gene expression at the transcriptional or post-transcriptional levels. But, the two types
of models discussed above for the ovalbumin gene lead to different predictions ...
(Breathnache/a/. 1977)
Such a model predicts that many cellular mRNAs derived from hnRNA will contain
spliced sequences and that all the sequences in such mRNAs were once part of a single
initial transcript. With the development of DNA recombinant techniques, such
suggestions should be testable. (Berget et al. 1977a)
Such a mechanism immediately raises new possibilities in interpreting physiological
and genetic data ... Fortunately, the technology seems advanced to a state which
should soon permit us to test these interesting possibilities. (Leder et al. 1977)

Such references to testability show the writers' deference before the community
of researchers (and also gives these researchers something to do).
The strategic use of pessimism is another way of demonstrating proper
deference to the community. Actually, one can only be pessimistic about fairly
strong claims; the assumption is that if one really thinks the claim won't be borne
out, one shouldn't publish it. So when Watson and Crick qualify their claim at
the end of their first article, it does not suggest that they thought their structure
unlikely to work out.

So far as we can tell, it is roughly compatible with the experimental data, but it must be
regarded as unproved until it has been checked against more exact results. (Watson and
Crick 1953)

Cech's group makes a similar remark in proposing a structure for a Tetra-

hymena intron:
Although the proposed structure is consistent with the nuclease digestion data, it
cannot be considered proven. Phylogenetic comparison ... will provide additional
evidence... (Cech etal. 1983)
20 THE PRAGMATICS OF POLITENESS

Such pessimism, like the invocation of testability, not only leaves the claim to the
reader's judgement, but also allows for further work by others, who can prove or
disprove the claim and share part of the credit.
Another device for redressing FTAs, less commonly used than hedging or
impersonal forms, involves acknowledging the FTA and apologizing for it. This
is effective, not because the apology in any way makes up for the FTA, but
because the apology shows that the writers recognize and accept the standards
of the community or the wants of the individual researchers who might be
slighted. Such apologies are common in redressing the relatively minor FTA of
omitting references from a review article. For instance, Francis Crick ends his
review with a note apologizing to researchers working with certain kinds of
organisms:
References have been kept to a bare minimum. I hope those working on mammalian
viruses and the immune system will forgive me for not describing their results more
fully. To do so would have made the article far too long. (Crick 1979)
In this case the FTA may be more significant than it seems, for Crick brushes off
the general model of 'multiple-choice genes' proposed by some researchers in
these areas, and says that they are dealing with 'a special case'. One reason
outright apology is a relatively rare device may be that it requires reasoning
about the personal responses of the writers and readers (readers grant that the
writers are aware of the offence to the readers) which is itself a threat to the
readers. Brown and Levinson can call on this sort of mutual awareness
gracefully within the discourse of linguists;
With some temerity we suggest that this simple reasoning might be applied quite
generally to the analysis of rituals of exchange. (Brown and Levinson 1987:46)
But in the discourse of molecular biology such remarks are made only by those
writers who, like Crick, can use their personae as part of their scientific
presentation.

BALD-ON-RECORD
Brown and Levinson outline a whole hierarchy of politeness strategies, and
argue that cultures can be compared in terms of which categories of redress are
preferred. Most of the claims and denials of claims in this corpus of scientific
articles illustrate either positive or negative politeness strategies. But there are
instances that illustrate the other choices open to Brown and Levinson's Model
Person: doing the FTA without redress, baldly; doing it off the record,
indirectly, and deciding not do do it at all. Their model can serve as a heuristic to
show us politeness where we might not expect to find it. Attention to the
circumstances in which writers choose these other strategies can allow us to test
some of the assumptions we have made about the culture of science.
Given the conventional notion that scientific articles are impersonal and
factual, one might expect that many face threatening acts would be done without
any redress, in the manner Brown and Levinson term 'bald-on-record'. But bald
 GREG MYERS 21

statements of claims, or criticisms ofrivalsare the exception, so much so that, as

we have seen, if an important statement seems to be made without modification,
one can probably assume that it is not the main claim of the article, not the
statement that is to be taken as new knowledge.
We do find the 'bald-on-record' strategy when, as Brown and Levinson put it,
'S wants to do the FTA with maximum efficiency more than he wants to satisfy
H's face ...' (page 95). In scientific articles, this means we find it where the
imposition on the reader is so small it can be ignored, or where the demands for
efficiency are so great that they override the demands of politeness. So, for
example, in announcing a big discovery that makes a small imposition, the
opening claim of 'A new method for sequencing DNA' is done baldly:
DNA can be sequenced by a chemical procedure that breaks a terminally-labelled
DNA molecule partially at each repetition of a base. (Maxam and Gilbert 1977)
Here there is no hedging, no device like 'We wish to suggest' or 'Results suggest
that DNA can be tested'. But this omission is possible because of the nature of
the FTA committed. It is not that the claim is small or affects few other
researchers: Maxam and Gilbert are announcing the technique that has since
been used by thousands of laboratories, has revolutionized molecular biology,
and has won Gilbert the Nobel prize. The announcement of the technique could
be said to threaten the negative face of all their rivals, the esoteric audience of
researchers who might have hoped to make such a breakthrough first, but this
threat is outweighed by the assumed benefit to most readers. For the same
reasons, Maxam and Gilbert, and any article giving methods, can use cookbook-
like imperatives throughout the article without the threat that Brown and
Levinson associate with imperatives. The reader who reads the command:
Dissolve dephosphorylated DNA in 75 microlitres of 10 mM glycine-NaOH...
is more interested in learning the method than in protecting his or her freedom
of action in regard to dissolving DNA. It is the equivalent of such politely bald
imperatives as 'Come in' or 'Take a seat'.
The other situation in which one might find a bald-on-record strategy is
where demands of efficiency—and especially brevity—overrule other con-
siderations, as in abstracts or in the very short, rapidly published articles known
as 'letters'. Watson and Crick's 1953 article in Nature is one of these letters, and
its abrupt dismissal of Fraser's model10 contrasts with the more detailed reason
given in a later, longer article in the Proceedings of the Royal Society (Watson
and Crick 1954). In some of the articles in the corpus drawn upon here the claim
is given baldly in the abstract, and with some modification in the text. This is the
first sentence of an abstract:
The ovalbumin gene is split in chicken DNA. (Breathnach et al. 1977)
Later in the article, the claim is stated with a note of personal response that, as
we have seen, mitigates the FTA.
22 THE PRAGMATICS OF POLITENESS

Unexpectedly, we have found that the DNA sequences complementary to ovalbumin

mRNA are split into several fragments in oviduct DNA... (op. cit.)
What is surprising is that the baldness of methods sections, abstracts, and some
letters is not possible in most scientific writing. But this is consistent with our
assumption of a great social distance between participants: only the smallest
FTA can be done baldly without risk.
OFF-RECORD
One might not expect to find in scientific articles any examples of 'off-record'
FTAs, in which the FTA is made only by implication, not in the literal sense of
the statement. The purpose of these articles is to record claims, and an explicit
statement is usually required to establish priority. One example of hinting at a
claim might be the famous sentence near the end of Watson and Crick's article,
where they say:
It has not escaped our notice that the specific pairing we have postulated immediately
suggests a possible copying mechanism for the genetic material. (Watson and Crick
1953)
The peculiar form in which the claim is put could be seen as the result of
politeness considerations. They cannot go into the genetic implications in a
short article, but it is politely assumed that any reader would see these
implications as Watson and Crick saw them without their spelling the implica-
tions out in the article. The phrase 'It has not escaped our notice' nods in the
direction of the reader, assuming that writers and readers share the same
powers of reasoning on this point. Crick said of this sentence later, 'In short, it
was a claim to priority' (Crick 1974). Such a claim made baldly ('The base
pairing is the mechanism of genetic copying') would imply that the readers were
competitive, and that this was a matter worth competing about, and would thus
be impolite. This phrasing takes it as an obvious implication, with no need to
claim priority. The indirectness seems to have annoyed some readers, as Crick
said later. 'This has been described as "coy", a word that few would normally
associate with either of the authors, at least in their scientific work. In fact it was
a compromise' (Crick 1974). Between Watson's desire not to mention it at all,
and Crick's to stake a claim, they arrived at an off-record claim. But claims can
be made in this indirect way only in unusual circumstances in which careful
attention to the article is guaranteed, so that no implication, however indirect,
will be missed. Otherwise a rival formulation is likely to be preferred. Coyness
has its dangers.
It is more likely that we will find such off-record FTAs with criticism of other
scientists or negative comments. I can given an example from a biological
controversy (not in molecular genetics) involving the behaviour of lizards.
Michael Moore, David Crews, and their colleagues respond to critics of their
work, and then say,
Gustafson and Crews (1981) have demonstrated experimentally that the presence and
behaviour of cagemates causes captive C. uniparens to produce more clutches of eggs.
 GREG MYERS 23

An understanding of the obviously complex social biology of unisexual Cnemidopho-

rus will be advanced only byrigoroustesting of hypotheses. (Moore et al. 1985)
The second sentence violates Grice's Maxim of Quantity, since both writers and
readers assume that science works by rigorous testing of hypotheses. So the
reader must look for some assertion implicated by this apparently redundant
statement. The reader sees from the previous sentence that Crews's group does
test hypotheses experimentally. So it is implicated that other researchers on
Cnemidophorus are not testing hypotheses. So they must be dismissing
hypotheses without scientific consideration. To say this directly would amount
to a personal insult.
The molecular biologists I am studying do not usually participate in the
vitriolic exchanges of the Cnemidophorus workers, but there are cases in which
negative comments are left to implicarures. Thus, commonly, a criticism will be
implicated only by the juxtaposition of statements that might at first seem
unrelated. The reader constructs a logical relation between them, the unstated
conclusion of which is that so-and-so is wrong. One example is in the full caption
and figure from one of Blake's articles.
Fig. 1. Plot of the inter-residue distances for each residue of bovine carboxy-peptidase.
This is equivalent to a "Go plot" in which the "valleys" along which the splice junctions
are claimed to run correspond to minima in the distance sum. The splice junctions in
the rat carboxypeptidase gene6 are marked by broken lines. The locations of the eight
beta-strands in the central beta-sheet of the carboxypeptidase molecule are shown.
(Blake 1983)

I have already quoted the second sentence as an example of a suggestion of

doubt mitigated by a passive construction and the use of a personal name only as
a term. The third sentence, taken literally, just draws attention to the lines
marking the splice junctions in the figure. But if one looks at the figure, one sees
that they do not fall on the minima, and that therefore the Go claim is wrong.
The fourth sentence draws attention to another way of grouping the residues,
and if one looks at the figure, one sees these marks do correspond with the
minima. Thus Blake has implicated that Go is wrong and that his explanation fits
her own graph better—but he has not made any explicit criticism.
Since quotation is so rare in scientific texts, any echoic language is likely to be
taken as ironic, as disavowed by the writers, even if it is just a word or two.
Doolittle makes such an off-record dismissal when he presents the consensus
view of evolution:
Most molecular biologists assume that the eukaryotic genome arose from a "simpler"
and more "primitive" prokaryotic genome rather like that of Eschaerichia coli.
(Doolittle 1978)
In a later article, Doolittle and Sapienza use an off-record strategy in question-
ing what they call the 'phenotype paradigm'. The questioning of an assumption
underlying much of the revolutionary biology and genetics, part of a neo-
Darwinian synthesis that has held (though not without criticism) since the
24 THE PRAGMATICS OF POLITENESS

1940s, is a very large imposition. But they are criticizing, not just the assump-
tions of their colleagues, but their manner of reasoning, their unscientific
attempt to save their neo-Darwinian assumptions at all costs. As we have seen,
one can always deny the claims of other researchers, but it is another matter to
question their motivations, their reasoning, their openness to a new paradigm:
that is to attack them as scientists, and it is an act one cannot do explicitly. So any
such criticism is entirely implicit in the passage as a whole.
The phenotype paradigm underlies attempts to explain genome structure. There is a
hierarchy of types of explanations we use to rationalize, in neo-Darwinian terms, DN A
sequences which do not code for protein. Untranslated messenger RNA sequences
which procede, follow, or interrupt protein-coding sequences are often assigned a
phenotypic role in regulating messenger RNA maturation, transport, or translation5"7.
Portions of transcripts discarded in processing are considered to be required for
processing8. Non-transcribed DNA, and in particular repetitive sequences, are thought
of as regulatory or somehow essential to chromosome structure or pairing. When all
attempts to assign to a given sequence or class of DNA functions of immediate
phenotypic benefit to the organism fail, we resort to evolutionary explanations. The
DNA is there because it facilitates genetic rearrangements which increase evolutionary
versatility (and hence long-range phenotypic benefit) 12"17, or because it is a repository
from which new functional sequences can be recruited18"19 or, at worst, because it is the
yet-to-be eliminated by-product of past chromosomal rearrangements of evolutionary
significance919. (Doolittle and Sapienza 1980)

Any reader can tell that they think this is a paradigm ready to crumble, but there
seem to be no explicit statements of doubt, and indeed they say 'most of us' are
'wedded' to this paradigm and 'such interpretations of DNA structure are very
often demonstrably correct'. There are a few implicit signals that the paradigm is
indeed being challenged here, such as the use of the term paradigm, which, by
invoking Thomas Kuhn, suggests the possibility of a coming paradigm shift.
Also, there are several phrases that suggest an attempt by biologists to save the
paradigm at all costs (rationalize; when all attempts fail; somehow essential; we
resort to evolutionary explanations). And the persistently impersonal phrasing,
which as we have seen is often used to mitigate criticism, prepares the readers
for a denial of these views.
But the key to our doubt is the continued focus on the activity of explanation
rather than the explanation itself, the piling up of defences of the paradigm given
briefly with echoes of the defenders' statements. Nearly every sentence is stated
so as to suggest a view rather than a fact (are often assigned; are considered; are
thought of). By the last sentence of the paragraph, readers have so thoroughly
suspended their belief that the sentence need not be hedged at all, and is still
read as making a questionable statement. It is the cumulative effect of the echoic
references to so many defences that makes us question them all as being ad hoc.
Following the reasoning of Sperber and Wilson in their presentation of echoic
utterances (1986:238-42), a reader might interpret the passage in the following
way: 'This is confusing, because there are so many different possibilities. The
language of these arguments is echoed but the arguments themselves are not
 GREG MYERS 25

developed, suggesting that the writers distance themselves from these state-
ments. But one assumes that Doolittle and Sapienza are using the most relevant
stimulus. They present the defences of the paradigm, each perhaps believable in
itself, with a number of others. Therefore they must mean for me to see this as an
uncontrolled proliferation of defences. Thus they are saying that the defences of
the paradigm must be questioned.' If the reader's reasoning is anything like this,
then Doolittle and Sapienza have made their criticism indirectly, through
inferences the reader makes from the text, and especially through the use of
echoic language.

AVOIDING THE FTA

The last option in Brown and Levinson's hierarchy of strategies is simply not to
do the FTA. This is a rational choice for those FTAs for which the rank of the
imposition is too great, or the power of the offended party too great, or the social
distance is too great, for any politeness device to mitigate the offence. Of course,
there is no explicit record where Speakers have decided not to say something, so
Brown and Levinson do not develop this category any further. (If there is the
documentable implication that the Speaker is withholding an FTA, then the
instance illustrates an off-record strategy, not a decision to avoid the FTA). So
for my examples of not doing an FTA I have to turn to another corpus of texts,
where I have access to comments by readers and revisions by writers in texts
that are part of a controversy.11
The FTA here is much more threatening than that made by Doolittle in the
previous example. It is a direct attack on the procedures of specific researchers;
the writer has made a new and startling claim, and two other researchers have
criticized his inferences and his technical skills. He, in turn, raises the issue of
their excessive commitment to established ideas, which make them incapable of
proper observation. All of this is clear enough from correspondence and
interviews, but such charges cannot be made explicitly in print. One can always
say that so-and-so's group have failed to take into account such-and-such, or
that their work is inconsistent with such-and-such evidence, or that their
procedures produce an artifact. But one cannot say that they are incompetent,
that they are unscientific, that their motives are mixed, just as one cannot, in the
House of Commons, call another MP a liar. There is a functional reason for this:
if one does not act as if one trusted all other scientists to be scientific, the whole
elaborate system of knowledge production would come tumbling down. But
there is also a rhetorical reason for avoiding personal criticisms: any personal
remark about another scientist is itself a sign of personal involvement on the
part of the critical scientist. So all personal criticisms must either be posed in
acceptable, impersonal, technical terms, or avoided.
For instance, in a manuscript, David Crews makes a very loaded criticism of
the work of other researchers: he criticizes them for basing their claims on
unrepresentative animals studied under 'artificial' conditions.
Much of the information on the causal mechanisms of vertebrate reproductive
behavior has been gathered on highly inbred stocks of rodents and birds living under
26 THE PRAGMATICS OF POLITENESS
artificial conditions... Some of the organismal concepts that have emerged are overly
narrow and sometimes unrealistic.
The comment could hardly avoid antagonizing the majority of readers: such a
criticism of the choice of research animal threatens to undermine other
researchers' whole careers, spent on elaborate experiments with hundreds of
mice or gerbils. After several reviews and several revisions, this opening is toned
down so that any suggestion of inbreeding, artificiality, narrowness, and
unreality is replaced with an uncontroversial comment on the kind of animals
studied.
A common observation of seasonally breeding vertebrates is that the reproductive
process of gamete production, sex hormone secretion, and mating behavior coincide,
and further, that sex steroid hormones activate mating behavior. The postulate of
hormone dependence of mating behavior is based primarily on detailed studies of
laboratory and domesticated species. (Crews 1984)
It might be argued that this says the same thing as the manuscript version. But
the view to be attacked is here given the benefit of consensus ('a common
observation') and the methods of researchers are praised ('detailed studies'),
and their selection of 'laboratory' species seems to be, at least at first, a
reasonable enough choice. There is no suggestion either that the researchers are
personally committed to their inbred mice and gerbils, or that Crews is
undertaking a personal crusade against them.
There are many such decisions about redress or avoidance of FTAs made in
any piece of writing, but, in academic writing, writers need not make such
decisions themselves. There is usually a whole series of readers, referees, and
editors who make comments and suggestions about what the responses of
readers of the published text will be. These comments show that scientists, as
readers and writers, are very sensitive to FTAs such as those I have discussed.
For instance, in another manuscript, Crews and his colleague Michael Moore
have a sentence that questions the competence of other (more experienced)
researchers on the Cnemidophorus lizard to observe some aspects of its
behaviour.

However, these investigators were not trained behaviourists and so apparently did not
recognize the unique opportunity this phenomenon presented as a model for under-
standing the functional linkage between sex and sexuality.

Two referees who give generally favourable and even enthusiastic reports on the
paper question the tone of this sentence, marking it for deletion and com-
menting ironically by echoing the phrasing. One says: 'It would at least be fair to
other researchers (whether or not they are "trained behaviourists") to state that
the behaviour that is the issue in the controversy had only been seen in captivity.'
Another referee comments on a technical point, made later in the paper, as
follows: 'The authors as "trained behaviourists" should know that social stress is
a complex phenomenon . . . ' Crews and Moore took the point of these ironic
 GREG MYERS 27

echoes and replaced the offending sentence with one that does not imply a
criticism.
Behavior patterns resembling courtship and copulation in all-female whiptail lizards
had not gone unnoticed by previous researchers (e.g. Schall, 1976; Cole and
Townsend, 1983; Cuellar, 1981). For the behavioural biologist, these observations
immediately raised a number of questions. (Crews and Moore (in press))
This version also points out the difference between Crews's project and that of
other researchers, but stresses the communality of the project. The other
researchers might object vigorously to being incorporated into Crews's project
this way. But the purpose of avoiding the FTA here is not to avoid offending
individual researchers, but to avoid offending the exoteric audience that might
see the whole affair as a personal squabble. The avoidance of the FTA saves the
writers' face by showing them to be apparently disinterested in their assessment
of the literature.
Some other referees' comments on manuscript of Crews's and Moore's paper
show an awareness of some of the other politeness strategies I have discussed. A
referee has written '?enough qualification??' and circled the hedging words in
the sentence:
The possibility that male-like courtship and copulatory behavior is hormonally
activated in an all-female species potentially represents the evolution of a novel
hormone-brain-behavior relationship.
Here the redundancy of hedging suggests, not politeness, but vagueness. The
same critic writes 'rather rhetorical and unnecessary' after a rhetorical question
on the next page:
If androgens are not male-like courtship and copulatory behavior, what is?
The rhetorical question is itself a politeness device, a case of an author
redressing the FTA of criticizing his or her critics. But in this form it is too
obviously personal, recognizably 'rhetorical'. Finally, the referee has marked for
deletion a particularly acerbic response to critics in the conclusion, which ends
with the sentence:
Obviously, if we are able to understand the complex social biology of unisexual
Cnemidophorus, we must rely less on polemics and more on rigorous tests of these and
other hypotheses.
Again, the FTA here is redressed with the expected politeness devices, a
reference to a common project, a first person plural pronoun identifying with
the field as a whole, hedging modifiers (less and more) and a general rule
(testing hypotheses). But to the referee it is still unacceptably personal and rude,
and he or she simply underlines the words, less on polemics and comments
'yes!'. Other readers, too, would presumably respond to the interactions
suggested in these statements, and see the underlying personal dispute.
28 THE PRAGMATICS OF POLITENESS

ANOTHER GENRE: POLITENESS IN POPULARIZATIONS

Brown and Levinson say that their hierarchy of politeness strategies can be
tested by comparing the strategies used in situations of varying social distance or
varying relative power of the speaker and the hearer. But in scientific articles, as
I have noted, these variables are assumed to be fixed, with very great distance
between the participants, and with no difference in power between those
individuals who are allowed to publish at all, but a great difference in power
between any individual and the scientific community as a whole. We can get
some confirmation that these various features are indeed the result of the
relations between writers and readers by comparing them to features in other
genres of writing on the same topic, such as news reports, popularizations,
textbooks, and personal accounts. These genres have the same topics, and often
the same authors, as the articles in the scientific journals, but the reader/writer
configuration is changed. Popularizations written by the researchers themselves
(like the articles in Scientific American) present many of the same problems of
FTAs as articles for scientific journals, but there is no longer the need to make
and deny claims. Popularizations and textbooks where the writer is not involved
in the research described pose different problems. In contrast to the double
writer and double reader configuration I proposed earlier, these articles put the
writer between the researchers and the readers.
Researchers Writer Readers
(esoteric group) (exoteric audience)
The writers of these popularizations (like, for instance, many of those in New
Scientist) do not have any tensions between their roles as writers and their roles
as researchers, and they only have to think of the general audience, not of the
other researchers in the esoteric group. (These other researchers would be
likely to read the article, but again, we don't need them to account for any
features of the text). It is the face of the exoteric group that is to be primarily
considered, not that of the esoteric group of researchers. The main interactive
problem here is to avoid insulting the readers, to try to make them feel like part
of the community of molecular geneticists, while still maintaining the proper
deference towards the scientific community.
This different configuration of readers and writers makes for different
politeness strategies. For instance, an unsigned article in New Scientist,
reporting on one of the articles in the corpus, gives an unhedged version of the
claim and a personal attribution of the results.
The immunologjcal landmark of the year is without any doubt the cloning and
sequencing for the first time of chromosomal DNA coding for an antibody. Susumu
Tonegawa and his colleagues at the Basle Institute of Immunology were the ones to do
it, with Wally Gilbert and his colleagues at Harvard, and what they have now is the
complete sequence of a fragment of mouse embryo DNA containing the antibody
variable region gene (see top diagram). (New Scientist 1977)
 GREG MYERS 29

It is acceptable for an outsider to praise this work, because the writer is just
conveying the credit already accorded by the community. It is acceptable for an
outsider to take a personal interest in the scientists, without weakening their
claims, because the main problem here is to make a link between these
researchers and the exoteric audience. The personal interest is emphasized by
the nickname (Wally), the listing of institutions, and the syntax. The sentence is
not
Immunoglobin wasfirstcloned and sequenced by Tonegawa,...
but
Tonegawa and his colleagues ... were the ones to do it...
This sentence construction treats the sequencing as a set task (do it), a prize that
these people have won. Just as these news reports need not hedge claims, they
need not be tactful in dismissing views now rejected by the esoteric group. Roger
Lewin is blunt in another article in New Scientist.
Some of the handwaving explanations of this apparent genetic profligacy have invoked
"extra genetic elements needed for controlling gene expression" and "essential
structural features required for packing the DNA in chromosomes." (Lewin 1979)
The modifier handwaving, drawn from the jargon of informal scientific discus-
sion, tells us these views are wrong, and the echoic statements bring out the
irony. The only mitigation is the impersonal form of the statement; these
quotations are not attributed to anyone.
Textbooks have a similar relation of writers to readers, but the readers now
stand in, as it were, for the researchers themselves, so the researchers hardly
figure at all. They can be as heavily hedged as scientific articles, but here the
hedging does correspond to the probability of the statement, instead of
reflecting the weightiness of the FTA of the claim. They are not as personal as
most popularizations, and indeed many textbooks hardly mention human
agents at all.
What kind of signal might we expect for precise intron excision in virtually every
eukaryotic messenger transcript? It is unlikely that each gene has its own unique
signal—far more likely that some ubiquitous sequence is present at exon-intron
junctions in all or most cases, because excision and splicing are ubiquitous processes.
By comparing the nucleotide sequence of mRNA with the sequence of the intact gene,
we find that no extensive sequence homology exists. Instead, very short sequences are
found in virtually every exon-intron junction that has been examined. These conserved
sequences are called consensus sequences, and they are characterized at the least by
the presence of a particular nucleotide doublet at each end of an intron. (Avers 1984)
There is no reference here to Chambon and Breathnach, who are credited with
first drawing attention to consensus sequences. Instead, we do everything, or
actions take place in passive sentences. There is hedging, but terminology and
claims are not themselves FTAs to be redressed. Undergraduate students
whose only experience of scientific writing comes from textbooks are not likely
30 THE PRAGMATICS OF POLITENESS

to have a sense of the social interactions involved in making claims, naming, or

speculating, and in criticizing, denying, or ignoring the claims of others.
Textbooks are scientific writing without the politeness.12

CONCLUSION
I have argued that the basic framework for the analysis of politeness can be
extended to written texts, if one can analyze the relations of writers and readers
instead of assuming a simple two-sided Speaker/Hearer relation. I have found
written examples to illustrate many of the categories of Brown and Levinson
devised for spoken interactions. But my purpose in making this argument is not
just to provide more evidence for the universality of politness strategies. Rather
I see it as a special case in the problem of studying the pragmatics of written texts
within a discourse community, a case that can tell us something both about
writing and about science. It is important for discourse analysis and for the
teaching of writing to show that, while writing does not involve face to face
contact, it is a form of interaction. Some aspects of this interaction are just what
we would expect from reading Brown and Levinson, Lakoff, and Leech on
politeness; for instance, writers use pronouns and other devices to stress
solidarity with readers, or use various hedges to modify statements that could be
FTAs. But this particular case also brings about some unexpected features, such
as the elaborate forms for suggesting shared assumptions even in criticisms (in a
culture stressing deference to the community), and the use of personal attribu-
tion to hedge claims or to make criticism (in a culture in which only claims
accepted by the consensus have authority). I have also looked at some of the
relatively rare examples of off-record and bald-on-record FTAs because these
unusual examples can tell us about the usual assumptions concerning relations
between writers and readers. Finally, I have briefly mentioned another genre to
which these articles could be compared, demonstrating that the kinds of devices
used in scientific articles are determined by the particular configuration of
writers and readers involved.
I also see this case as a way of using an established method—the analysis of
politeness strategies—as an aid to the study of a culture that is not usually seen in
these ethnographic terms. For instance, the choices of politeness strategies
would suggest that claims and denials are indeed taken as possible impositions
on the community, that the coining of new terms is also a possible imposition,
that speculation must be done but must be apologized for, that hedging reflects
social interactions rather than probability of statements, and that direct
criticism is almost inadmissible. Politeness theory is useful to sociologists of
science in shifting attention from static rules, norms, and systems to the
dynamics of interactions. This is a useful reminder to those of us who study
scientists, because it suggests we look for explanations of the features of
scientific writing, not in any presupposed universality, communality, scepticism,
or disinterestedness following from the scientific method, but in the daily
interactions of competing scientists.13
(Received October 1987)
 GREG MYERS 31

NOTES
1
Analyses from conversational pragmatics have served as guides for other analysts of
texts: see, for example, Widdowson (1979) and Hoey (1983).
2
Erving Goffman (1981) has written on similar problems with spoken interactions in
his chapter on 'Footing'.
3
For instance, when Brown and Levinson dismiss Geoffrey Leech's Maxims of
Politeness, they commit an FTA in relation to a particular researcher, Leech, but not in
relation to me, a member of the wider audience. They do not actually address Leech. But I
will expect them to mitigate this FTA by showing some respect for Leech's face. On the
other hand, if Brown and Levinson were to claim credit from the linguistic community on
the basis of some as yet unsupported speculations, they would be committing an FTA
with regard to the exoteric audience, but not with regard to Geoffrey Leech, Robin
Lakoff, or other members of the esoteric group of researchers on politeness.
4
The most influential analysis of the ethos of science, that of Robert Merton (1942;
reprinted 1973), defines four norms essential to the function of science as a system:
universalism, communism, organized scepticism, and disinterestedness. The devices I
find are consistent with such an analysis, but following Brown and Levinson's criticism of
approaches based on norms, I think that these devices are better explained in terms of
rational strategies. For an introduction to the sociological critique of the Mertonian ethos
of science, see Mulkay (1979).
5
In addition to these threats to the readers' negative and positive face, there are a
number of possible threats to the writer's own positive or negative face, involving
violation of the Mertonian norms: to offer or accept personal praise goes against
universalism, to withhold information goes against communism, to express emotional
commitment violates disinterestedness, and to express personal, unsupported belief
violates organized skepticism. Such acts are common enough in less formal discourse, in
scientific autobiographies, or popularizations, or conversation, but they are uncommon
in scientific articles, so I will not be considering them here.
6
Brown and Levinson argue that strategies at one end of their hierarchy (towards
doing the FTA baldly or using positive politeness strategies) are preferred for less weighty
impositions, while strategies at the other end (towards not doing the FTA at all, doing it
off record, or using negative politeness strategies) are preferred for weightier impositions.
They also argue that in a culture with great social distance and small differences in
individual power (such as the culture of science) there will be a tendency to favour
negative politeness and off-record redress. Both these generalizations apply, more or less,
to my analysis of the kinds of politeness devices used in scientific articles.
7
Recall that the first of the studies on politeness was Robin Lakoff s paper for the
Chicago Linguistic Society entitled 'The Logic of Politeness; or, Minding Your p's and
q's'. In fact, punning titles are a characteristic feature of the Proceedings of the Chicago
Linguistic Society.
8
So, for instance, when I want to cite the literature on politeness, I could legitimately
refer only to the new edition of Brown and Levinson, since it has become the classic
statement and it refers to any other papers that might be relevant. But here is how
politeness enters in: when Stephen Levinson refers to the literature on politeness in his
textbook on pragmatics, he has to cite Robin Lakoff and Geoffrey Leech as well, even
though Brown and Levinson don't entirely agree with these authors, or rather, especially
because Brown and Levinson don't entirely agree with them.
Similarly, in molecular genetics, the only authors who have to keep citing all the rival
32 THE PRAGMATICS OF POLITENESS

claims are the authors of the heavily cited article; they have to demonstrate again and
again that they are aware of all the other groups working on the same problem. So, for
instance, an article by Carey's group came out at almost exactly the same time as an article
by Chambon's group, also finding the ovalbumin gene was split, but using somewhat
different methods, and making a rather more cautious interpretation of their findings. The
article was not much cited, perhaps because the claim was not made in terms of split genes
(they refer to the gene being found in a number of different restriction fragments, not to it
being split). But years later, Chambon was still citing it, and referring to the work in some
detail, whenever he gave an account of the work of his own group. An article by J. E.
Darnell put forward a hypothesis about evolution just a few months after a shorter and
sketchier article by W. F. Doolittle; the idea is usually referred to with a citation of
Doolittle, except by Doolittle himself, who also cites the slightly later article by Darnell.
9
John Sinclair has pointed out to me that the effect of such an attribution varies with its
place in the statement: when the attribution is inserted in the middle of the statement, as
the he claims is here, it is more likely to be a hedge, but when it comes at the beginning, it
may be taken as a point of information, and not weaken the statement.
10
'This structure as described is rather ill-defined, and for this reason we shall not
comment on it' (Watson and Crick 1953).
1
' For more on these cases, see Myers (1985a and b; forthcoming).
12
An awareness of politeness devices may also help teachers understand some aspects
of the style of university students who are native speakers of English. Generations of
freshman composition teachers in the United States have tried to make the writing of their
students conform to the prescriptions of such style manuals as Strunk and White or, more
recently, Joseph Williams, prescriptions warning the student against nominalization,
passives, and the 'wordiness' of hedging and indirect statements. Teachers have
sometimes interpreted the failure of their students to follow these prescriptions as further
evidence of the corruption caused by the prose of sociologists, politicians, and journalists.
But students' preference for passives and hedging could be seen, not as the imitation of
conventions, but as rational attempts to find the best strategy for interactions in academic
writing. A student who tries to mitigate the FTA of making claims will violate many of the
rules of composition textbooks.
13
Parts of this paper were given at the autumn 1987 meeting of the Linguistics
Association of Great Britain, at the English Language Research Seminar, University of
Birmingham. My thanks to those who made suggestions at these presentations, and to
Roz Ivanic, Tony Dudley-Evans, John Green, and Susan Price.

REFERENCES
Brown, P. and S. Levinson. 1987. Politeness: Some Universals in Language Usage.
(Reprinted, with a new introduction, from E. Goody (ed.): Questions and Politeness.)
Cambridge: Cambridge University Press.
Brown, R. and A. Gilman. 1960. 'Pronouns of power and solidarity' in T. A. Sebeok (ed.):
Style in Language. Cambridge, Mass.: MIT Press.
Fleck, L. 1935; 1981. The Genesis and Development of a Scientific Fact. (German edn.
1935.) T. Trenn (trans.), Chicago: University of Chicago Press.
Goffman, E. 1967. Interaction Ritual. Garden City, NY: Doubleday.
Goffman, E. 1981. Forms of Talk. Oxford: Basil Blackwell.
Hagstrom, W. O. 1965. The Scientific Community. New York: Basic Books.
Hoey, M. 1983. On the Surface of Discourse. London: Allen and Unwin.
 GREG MYERS 33

LakofT, R. 1973. 'The logic of politeness; or, Minding your p's and q's' in Papers from the
Ninth Regional Meeting of the Chicago Linguistic Society. Chicago: Chicago Linguistic
Society.
Latour, B. 1987. Science in Action. Milton Keynes: Open University Press.
Leech, G. 1983. The Principles of Pragmatics (2nd edn.) Harlow: Longman.
Merton, R. (1942) 1973. The Sociology of Science. Chicago: University of Chicago
Press.
Mulkay, M. 1979. Science and the Sociology of Knowledge. London: Allen and Unwin.
Myers, G. 1985a. 'The social construction of two biologists' proposals.' Written
Communication 2/3.
Myers, G. 1985b. 'Texts as knowledge claims: the social construction of two biology
articles.' Social Studies of Science 15/4.
Myers, G. (forthcoming). Writing Biology: Texts in the Construction of Scientific
Knowledge. Madison: University of Wisconsin Press.
Sperber, O. and D. Wilson. 1986. Relevance: Communication and Cognition. Oxford:
Blackwell.
Widdowson, H. G. 1979. Explorations in Applied Linguistics. Oxford: Oxford Univer-
sity Press.

SOURCES
Aloni, Y., R. Davi, O. Laub, M. Horowitz, and G. Khoury. 1977. 'Novel mechanism for
RNA maturation: The leader sequences of simian virus 40 are not transcribed adjacent
to the coding sequences.' Proceedings of the National Academy of Sciences 74/
9:3686-90.
Avers, C. 1984. Molecular Cell Biology. Boston: Addison Wesley.
Berget, S. M., A. J. Berk, T. Harrison, and P. A. Sharp. 1977a. 'Spliced segments at the
5' termini Adenovirus-2 late mRNA: A role for heterogeneous nuclear RNA in
mammalian cells.' Cold Spring Harbor Symposium on Quantitative Biology 42:
Chromatin. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
Berget, S. M., C. Moore, and P. A. Sharp. 1977b. 'Spliced segments at the 5' terminus
of adenovirus 2 late mRNA.' Proceedings of the National Academy of Sciences 74/
8:3171-5.
Blake, C. C. F. 1978. 'Do "genes-in-pieces" imply proteins in pieces?' Nature 273:267.
Blake, C. C. F. 1983. 'Exons—present from the beginning?' Nature 306:535-7.
Breathnach, R., C. Benoist, K. O'Hare, F. Gannon, and P. Chambon. 1978.
'Ovalbumin gene: Evidence for a leader sequence in mRNA and DNA sequences at
the exon-intron boundaries.' Proceedings of the National Academy of Sciences 75:
4853-7.
Breathnach, R., J. L. Mandel, and P. Chambon. 1977. 'Ovalbumin gene is split in
chicken DNA.' Nature 270:314-19.
Cech, T. R., N. K. Tanner, I. Tinoco, B. R. Weir, M. Zuker, and P. S. Perlman. 1983.
'Secondary structure of the Tetrahymena ribosomal RNA intervening sequence:
Structural homology with fungal mitochondrial intervening sequences.' Proceedings of
the National Academy of Sciences 80:3903-7.
Chambon, P. 1978. 'Summary: The molecular biology of the eukaryotic genome is
coming of age.' Cold Spring Harbor Symposium on Quantitative Biology 42:
Chromatin. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
34 THE PRAGMATICS OF POLITENESS

Chow, L.T., R. E. Gelinas, T.R. Broker, and R.J. Roberts. 1977. An amazing
sequence arrangement at the 5' ends of Adenovirus 2 late messenger RNA.' Cell
12:1-8.
Crews D. 1984. 'Gamete production, sex hormone secretion, and mating behavior
uncoupled.' Hormones and Behaviour 18:22-8.
Crews, D. and M. C. Moore, (in press) 'Reproductive physiology of parthenogenetic
whiptail lizards' in J. W. Wright (ed.): Biology ofCnemidophorus lizards. Los Angeles,
Calif.: Los Angeles City Museum.
Crick, F. 1974. "The double helix: a personal view.' Nature 248:766-71.
Crick, F. 1979. 'Split genes and RNA splicing.' Science 204:264-71.
Darnell, J. E. 1978. 'Implications of RNA-RNA splicing in evolution of eukaryotic
cells.' Science 202:1257-60.
Doolittle, W. F. 1978. 'Genes in pieces: Were they ever together?' Nature 272:581-2.
Doolittle, W. F., and C. Sapienza. 1980. 'Selfish genes, the phenotype paradigm and
genome evolution.' Nature 284:601-3.
Gilbert, W. 1978. 'Why genes in pieces?' Nature 271:501.
Go, M. 1981. 'Correlation of DNA exonic regions with protein structural units in
haemoglobin.' Nature 291:90-2.
Hsu, M-T., and J. Ford. 1977. 'A novel sequence arrangement of SV 40 late RNA.' Cold
Spring Harbor Symposium on Quantitative Biology 42: Chromatin. Cold Spring
Harbor, NY: Cold Spring Harbor Laboratory.
Jacob, F., and J. Monod. 1961. 'Genetic regulatory mechanisms in the synthesis of
proteins.' Journal of Molecular Biology 3:318-56.
Jacq, C, J. Lazowska, and P. P. Slonimski. 1980. 'Sur un nouveau mecanisme de la
regulation de l'expression genetique.' Comptes Rendus de VAcademie des Sciences de
Paris 290:89-92.
Klessig, D. F. 1977. 'Two adenovirus mRNAs have a common 5' terminal leader
sequence encoded at least 10 kb upstream from their main coding regions.' Cell 14:9-
21.
Kruger, K., P. J. Grabowski, A. J. Zaug, J. Sands, D. E. Gottschling, and T. R. Cech.
1982. 'Self-splicing RNA: Autoexcision and autocyclizarion of the ribosomal RNA
intervening sequences of Tetrahymena.' Cell 31:147-57.
Leder, P., S. M. Tilghman, D. C. Tiemeier, F. J. Polsky, J. G. Seidman, M. H. Edgell,
L. W. Enquist, A. Leder, and B. Norman. 1977. 'The cloning of mouse globin and
surrounding gene sequences in bacteriophage lambda.' Cold Spring Harbor Sym-
posium on Quantitative Biology 42: Chromatin. Cold Spring Harbor, NY: Cold Spring
Harbor Laboratory.
Lerner, M. R., J.A. Boyle, S.M. Mount, S. L. Wolin, and J.A. Steitz. 1980. Are
snRNPs involved in splicing?' Nature 283:220-4.
Lewin, R. 1979. 'Why split genes?' New Scientist 10th May. pp. 452-3.
Maxam, A. M., and W. Gilbert. 1977. 'A new method for sequencing DNA.' Proceedings
of the National Academy of Sciences 74/2:560-4.
Moore, M. C, J. M. Whittier, A. J. Billy, and D. Crews. 1985. Male-like behaviour in
an all-female lizard: Relation to ovarian cycle.' Animal Behavior 33:284-9.
Orgel, L. E., and F. H. Crick. 1980. 'Selfish DNA: the ultimate parasite.' Nature
284:604-7.
Padgett, R. A., S. M. Mount, J. A. Steitz, and P. A. Sharp. 1983. 'Splicing of messenger
RNA precursors is inhibited by antisera to small nuclear ribonucleoprotein.' Cell
35:101-7.
 GREG MYERS 35

Saint Girons, H. 1981. 'Cycle annuel et survie de quelques viperes d'Europe: Influence
des temperatures exceptionellement elevees de l'annee 1976.' Vie et Milieu 31/1.
Slonimski, P. P. 1980. 'Elements hypothetiques de l'expression des genes morceles:
proteines mesageres de la membrane nucleaire.' Comptes Rendus de VAcademie des
Sciences de Paris 290:331-4.
Tilghman, S. M., P. J. Curtis, D. C. Tiemeier, P. Leder, and C. Weissmann. 1978. The
intervening sequence of a mouse beta-globin gene is transcribed within the 15S beta-
globin mRNA precursor.' Proceedings of the National Academy of Sciences 75/3.
Tonegawa, S., A. M. Maxam, R. Tizard, O. Bernard, and W. Gilbert. 1978. 'Sequence
of a mouse germ-line gene for a variable region of an immunoglobin light chain.'
Proceedings of the National Academy of Sciences 75/3.
Watson, J., and F. H. C. Crick. 1953. 'A structure for deoxyribose nucleic acid.' Nature
171:737.
Weber, J., W. Jelinek, and J. E. Darnell, Jr. 1977. 'The definition of a large viral
transcription unit late in Ad2 infection of HeLa cells: Mapping of nascent RNA
molecules labeled in isolated nuclei.' Cell 10:611-16.