Toury, Gideon 1998. "A Handful of Paragraphs on 'Translation' and 'Norms'"
In: Christina Schäffner, ed. Translation and Norms. Clevedon etc.:
Multilingual Matters, 1998. 10-32. [also available as Vol 5, Nos 1&2 of
Current Issues in Language & Society
© All rights reserved.
Text scanned for educational use, Unit for Culture Research, Tel Aviv University

A Handful of Paragraphs on 'Translation'
and 'Norms'

Gideon Toury

The M. Bernstein Chair of Translation Theory, Tel Aviv University, Tel Aviv 69978,

The format of paragraphs has been chosen to present questions and a few tentative
answers on the theme of translation and norms. The formulation of questions is an
important aspect of any research programme, and it has been the basis for descriptive-
explanatory research as well. Translating as an act and as an event is characterised by
variability, it is historically, socially and culturally determined, in short, norm-gov-
erned. In the paragraphs below, the following issues are discussed: the relationships
between social agreements, conventions, and norms; translational norms, acts of trans-
lation and translation events, norms and values, norms for translated texts vs. norms
for non-translated texts, competing norms. Comments on the reactions to three differ-
ent Hebrew translations of Hemingway's short story 'The Killers' are presented at the
end of the paper.

1 An Introductory Note on Aims and Strategy of Presentation1

This text should not be regarded as a full-fledged paper which offers a
well-rounded presentation of all that may be invoked by the two title notions and
their possible combinations. The main aim of the text is to supply food for thought
for anyone wishing to get into the right mood and prepare for the Aston Seminar
on 'Translation and Norms'. Above all, it is meant to lay down some ground rules
for an open discussion.

As experts on diet know only too well, food is much more digestible when
served in small, well-dosed quantities. It is also much more appetising that way.
It is this kind of strategy that was adopted for the present document, and for the
very same purposes. Thus, my humble aim is to supply a cocktail (shaken, not
stirred) of select questions with a number of tentative answers and an odd (more
general) hypothesis. The document should therefore be seen as no more than a
series of paragraphs on the theme of our seminar; a (hopefully) coherent whole
with a minimum amount of cohesive devices, mainly a number of cross-refer-
ences, which should allow for a multi-directional kind of reading. I trust this
would be acceptable just the same, in view of the fact that, especially in the UK,
the notion of 'paragraph' has gained considerable circulation and a measure of
respectability as a mode of presenting ideas on translation.

In a way, this is a tribute to Peter Newmark, then, who invented the genre of
'Paragraphs on Translation'. However, as experts on multilingualism and
language transfer would surely appreciate, at the root of my thinking in terms of
paragraphs there was also contemporary colloquial Hebrew, where tshmá kéta
(literally: 'listen to a paragraph/segment' but actually something like 'wanna
hear something weird/funny') is a common discourse organiser; a marker of

1352 0520/98/01 001023 $10.00/0 c 1998 G. Toury

[pg.] 10

A Handful of Paragraphs on Translation and 'Norms'

[pg.] 11

opening, setting the tone for a rather casual mode of presentation (Maschler, in
press), which is only too fitting for the aims and strategy of my presentation.

Let me make the terms of reference of my position paper as clear as possible:
as always, my main interest lies with descriptive-explanatory research rather than
mere theorising. For me, theory formation within Translation Studies has never
been an end in itself. Its abject has always been to lay a sound basis and supply
an elaborate frame of reference for controllable studies into actual behaviour and
its results, and the ultimate test of theory is its capacity to do that service.

Moreover, I consider the formulation of questions an important aspect (and
phase) of any research programme. Therefore I truly believe it is questions we
should be focusing on in our seminar rather than any answers that may or may
not be suggested, this document included. At the same time, I hope we'll be
starting our negotiations with one agreement at least; namely, that the association
of 'translation' and 'norms' is not just valid, from the theoretical point of view,
but of potential value too, for whatever each one of us may be interested in doing
within Translation Studies; otherwise why take up this topic in the first place?
As will soon become clear, my choice of the word 'negotiations', in this
connection, is all but rhetorical.

In order to keep a minimum amount of order in this progression of
paragraphs, topics for possible discussion will be presented in two separate
clusters: a general one, tackling the notions of agreement, convention and norm
within a social setting, followed by a more specific cluster, where the same
notions will be taken up again and tied to translational behaviour. Some
methodological points will also be made where appropriate. Whatever is related
to these topic-areas will follow a second introductory section of a historical nature,
an attempt to supply some contextualising facts to the use of the notion of norms
in Translation Studies in the last few decades. Those who find this section too
long or too personal are advised to skip it and move directly to paragraph-cluster
3. I feel obliged to do it this way in view of the many distortions in the way recent
developments in the discipline have been presented, most notably in Gentzler
(1993), which seems to have become a standard work, in this respect.

2 Historical Observations on the Association of 'Norms' and

Let's agree to refrain from going into the question of who was the first to say
what. Due to our incomplete knowledge of the history of our own discipline,
where the wheel has been and is still being re-invented time and again, such
questions are bound to generate hot debates; which is not bad in itself, had it not
been for the fact that such debates would inevitably lead us way off track.
Personally, I am more than willing to waive all claims for originality of thinking
and give up any credit one might like to give me for having been the first to tackle
translation as the norm-governed behaviour it tends to be; credit which I never
claimed anyway.

Thus, it goes without saying that it wasn't I who suggested the association of
'translation' and 'norms'. This association was very much present, if only
implicitly, in the work of Jiri Levy, (1969 [1963]) and James S Holmes (1988), with
whom I have always felt the strongest affinity, as well as a number of other

[pg.] 12

Current Issues in Language and Society

scholars. All of these could easily have carried their research well into the realm
of translational norms because the foundations were certainly there. Needless to
say, they all had predecessors of their own, which would have made it possible
-- and not uninteresting -- to trace the association of 'translation' and 'norms'
further back.

All this notwithstanding, I am probably the one person who would have to
take the responsibility -- the blame, some will no doubt insist -- for having.
injected the heaviest dose of norms into the veins of Translation Studies in the
1970s and early 1980s, in as much as the substance thus injected indeed dissolved
into the bloodstream of the discipline (which is one thing I hope to see verified
during our seminar). At the very least I would have to be granted with having
made 'norms' a kind of legal tender in the discussion of translation practices and
their results; because there have surely been quite a number who have adopted
the term as little more than a catch-phrase. That is, without reflecting on any of
the necessary consequences, or even trying to find out what those consequences
might have been.

For me, the beginning was over 25 years ago, when I started researching for
my PhD dissertation; and the notion of norms first presented itself as a means of
elegantly bridging a gap I encountered while trying to account for the observed
results of translational behaviour during a limited period of time (the crucial
years between 1930 and 1945) in the history of translation of one text type, prose
fiction, within one culture/language: the Hebrew one.

The gap I am referring to was between the notion of translation as it had came
to be used by the beginning of the 1970s and the principles of establishing a
corpus for a descriptive-explanatory study such as the one I had in mind. The
main problem was how to draw a justifiable, non-arbitrary line between that
which would be included in the corpus because it pertains to translation as
conceived of by the culture in question, and that which would be left outside of
it because it does not. The necessary demarcation could simply not be worked
out on the basis of any of the conceptualisations I was able to lay my hands on,
and for quite a while I became a fervent collector of definitions of translation, in
the wild hope of hitting upon the ultimate one.

I soon realised that my difficulties stemmed from the very nature of the
essentialistic definition, imposing as it does a deductive mode of reasoning,
rather than the formulation of any single definition. Even the most flexible of
these definitions, as long as it still purported to list the necessary and sufficient
conditions for an entity to be regarded as translational, proved to be unworkable.
It then dawned on me that, in the very attempt to define translation, there was
an untenable pretence of fixing once and for all the boundaries of a category
which is characterised precisely by its variability: difference across cultures,
variation within a culture and change over time. Not only was the field of study
thus offered considerably shrunk, in comparison with what cultures had been
and were still willing to accept as translational, but research limited to such
pre-defined boundaries could not help but breed a circular kind of reasoning: to
the extent that the definition is taken seriously, whatever is tackled -- selected
for study because it is known to fall within its domain -- is bound to reaffirm it;
and if, for one reason or another, it is then found to be at odds with the initial

A Handful of Paragraphs on Translation and 'Norms'
[pg.] 13

definition, it will have to be banished from the corpus. In extreme cases, when
actual behaviour is in little congruence with the definition, there would remain
hardly anything to study as translations, which is inconsistent with presystematic
intuitions based on our acquaintance with the history of translation.

The way out of that deadlock seemed to me to try and have variability in all
its facets introduced into the notion of translation itself, whereby any kind of
realisation of that notion would necessarily be regarded as historically, socially
and culturally determined; in brief, as norm-governed.

Needless to say, any attempt to close the gap in any real manner necessitated
a lot more than the mere introduction of a WORD such as 'norm' into the theoretical
arsenal (in which Itamar Even-Zohar's 1971 seminal PhD dissertation was of
paramount importance, in my particular case). It had to be made operable. I
therefore invested time and effort in theoretical and methodological elaborations
on the NOTION of norm, especially in relation to its possible application to

The results came into the open during the pioneering Conference on Literature
and Translation, which was held in Leuven (Belgium) in 1976 (for the Proceedings
see Holmes et al., 1978) -- my first international conference ever. Unfortunately,
it was only the skeletal English version (Toury, 1978) rather than the full Hebrew
text (Toury, 1977) which became available, and that version was quite correctly
characterised as overly schematic; a reflection of its having been a mere summary.
In fact, schematism proved to be the strength of that paper as well as its weakness,
making it appealing to some and repellent to others, even to this very day. It
worked particularly well when regarded alongside Even-Zohar's presentation to
the same Conference, entitled 'The Position of Translated Literature within the
Literary Polysystem' (Even-Zohar, 1978). Unfortunately, the close connections
between the two were all too soon forgotten, which was no great help to the
appropriate reception of my first paper on norms.

Unlike the development in recent years, conferences on translation, especially
truly scholarly ones, were quite rare in the 1970s. Even against this backdrop, the
Leuven Meeting was a unique event: with very few exceptions, it brought
together a non-randomly selected group of relatively young scholars, many of
them graduate students like myself. I therefore found myself preaching to people
who were basically on the verge of conversion to a sociocultural way of thinking
about translation anyway. Many of them had, in fact, crossed the critical
threshold and were ready for more. This is probably the sense in which my
partner in the present Seminar, Theo Hermans (who was among the participants
of the Leuven Conference himself), later claimed that time was ripe for a change
of paradigms of this precise nature and that translation scholars were well
prepared for one brand or another of systemic reasoning which has the notion of
norms built into it (Hermans, 1995). Unfortunately, while this might have been
true for the group convening in Leuven and their disciples, it hardly held for
Translation Studies as a whole. I have already expressed my wonder as to the
extent to which it holds today...

At the beginning of the 1980s, a number of colleagues, especially younger ones,
adopted the notion of norms and tried to apply it to their own corpora, trying to
salve the problems they themselves were keen on solving. Some of them even

[pg.] 14
Current Issues in Language and Society

developed the notion itself in different ways, or at least criticised the simplicity
and rigidity of my skeletal presentation, making me rethink it in more and more
dynamic terms. Due to their personal backgrounds, most of the scholars who tried
their hands with the notion of norm were first and foremost engaged in the study
of literary translation. However, while this focused interest is easy enough to
explain, it is not the case that literature is the only domain where translation can
be expected to be norm-governed. It is simply that the notion of the norm has
hardly been put to a serious test as an explanatory tool in any other field. This,
in other words, is a weakness of Translation Studies in the present phase of its
evolution and of its proponents as individuals, rather than of the notion of the
norm itself, which has much wider, maybe even universal applicability (see first
attempts to apply the notion to Conference Interpreting, of all modes of
translation; most notably in Shlesinger, 1989 and Harris, 1990. To judge from
recent conferences, these attempts will soon be revived.)

This marks our transition to the cluster of paragraphs, beginning with the more
general ones.

3 Social Agreements, Conventions and Norms

Norm-hunting, in any domain of human behaviour, clearly indicates that a
sociocultural perspective was opted for. To the extent that such a perspective can
be justified, there is no escape from taking seriously what the Social Sciences have
to offer us. We need not become sociologists ourselves to do so. In our case, we
can still wish to salve the specific riddles of translation, but we would be doing
so on the assumption that translation is basically a sociocultural, and hence
norm-governed activity (see paragraph 4.1). In the following paragraphs, a
selection of sociocultural notions will be presented, in a way which would pave
the way for their subsequent association with translation.

3.1 Agreements and conventions

An important assumption of sociologists and social anthropologists is that
there must be some humanly innate flair for socialising, which some have named
sociability. This faculty is assumed to be activated whenever a number of persons
came into contact and start exploring their situation with a view to living
together. They do this whether what is at stake is the establishment of a new
group or just the sustenance of an existing one. As J. Davis -- an anthropologist
who tried to systematise the notion of 'social creativity' and render it serviceable
in explaining the making and maintenance of social groups -- recently put it:

People use their given sociability to create agreements about actions. So, our
worlds achieve the appearance of stability and regularity because we agree
that certain actions are acceptable in appropriate circumstances, and others
are not. (Davis, 1994: 97, italics added)

Rather than given, (tacit or explicit) 'agreements about actions' are always
negotiated, with or without the intervention of language. Such negotiations, which
require some time, result in the establishment of social conventions, according to
which members of the group will behave when they find themselves under
particular circumstances. They often do so in the form of behavioural routines.

A Handful of Paragraphs on Translation and 'Norms'

[pg.] 15

'What we create is -- within agreed limits -- a predictable event, from which
certain choices have been excluded'. 'So when we are [socially] creative we
attempt to create order and predictability and to eliminate choice, or at any rate
to confine choice within certain prescribed limits' (Davis, 1994: 97).

3.2 Negotiations and re-negotiations

In as much as a group is indeed formed, or its existence sustained, the process
involving negotiations, agreements and conventions-and-routines can thus be
regarded as inevitable. The exact way it proceeds in any individual case, by
contrast, is not given in any way. Rather, it is a function of the prevailing
circumstances. Many times it may even seem as if 'it could so easily have been
otherwise' (Davis, 1994: 97). At the same time, in retrospect, what was opted for
can normally be accounted for; the agreements themselves as well as the way
they were negotiated and reached.

Nor is the establishment of a societal group merely a time-consuming process.
In addition to its gradual nature it is also a never-ending one: as long as the group
has not collapsed, social order and everything that goes with it are constantly
being (re-)negotiated; the more so when new members wish to join the group or
when it is challenged by rivalling groups. Small wonder, then, that the process
also involves adjustments, and hence changes, of agreements, conventions and
behavioural routines. In fact, the most one would get is temporary, sometimes
-- i.e. in very unstable societies -- even momentary states of equilibrium.

3.3 Conventions and norms

Conventions are a necessary outcome of any striving for social order, as well
as a means for its attainment and maintenance. At the same time, they are not
specific and binding enough to serve as guidelines for (and/or a mechanism for
the assessment of) instances of behaviour and their products. Due to their
inherent vagueness, the acquisition of conventions poses special problems to
newcomers to a group, which would be the normal case with new translators
starting to work in and for an established society (paragraph 4.11). There is a
'missing link' here which the notion of norm seems a good candidate for

3.4 Norms

Norms have long been regarded as the translation of general values or ideas
shared by a group -- as to what is conventionally right and wrong, adequate and
inadequate -- into performance instructions appropriate for and applicable to
particular situations, specifying what is prescribed and forbidden, as well as what
is tolerated and permitted in a certain behavioural dimension (the famous 'square
of normativity', which has recently been elaborated on with specific regard to
translation, e.g. in De Geest, 1992: 38-40). They do so even if one refuses to accept
that values art as causal elements of culture, as a sort of ultimate ends towards
which action is directed, and maintains instead that

culture influences action not by providing the ultimate values toward
which action is oriented, but by shaping a repertoire or 'tool kit' of habits,

[pg.] 16

Current Issues in Language and Society

skills, and styles from which people construct 'strategies of action'.
(Swidler, 1986: 273)

As long as there is such a thing as appropriate vs. inappropriate behaviour
(according to an underlying set of agreements), there will be a need for
performance instructions as well. In a way, then, norms may be seen as part of
Swidler's 'tool kit'. while they may not be 'strategies of action' in themselves, they
certainly give rise -- and lend justification -- to such strategies.

3.5 Norms vs. normative formulations

Not only can social negotiations be carried out with or without the intervention
of language (paragraph 3.1), but also the norms themselves which would govern
behaviour need not be formulated at all: They may well remain implicit. At the
same time, there is always at least the possibility of having norms verbalised, in
order simply to comment on them (or on norm-governed behaviour and its
results) or even as part of the process of imparting them to others to ensure social

This possibility notwithstanding, it is important to bear in mind that there is
no identity between the norms as the guidelines, as which they act, and any
formulation given to them in language. Verbalisations obviously reflect awareness
of the existence of norms and their significance. However, they always embody
other interests too, particularly a desire to control behaviour -- i.e. dictate norms
(e.g. by culture planners) -- or account for them in a conscious, systematic way
(e.g. by scholars). Normative formulations may, therefore, serve as a source of
data on norm-governed behaviour, and hence on the underlying norms as such,
but they may do so only indirectly: if one wishes to expose the bare norms, any
given formulation will have to be stripped of the alien interests it has

3.6 Norms and regularities of behaviour

Obviously, there is a point in assuming the existence of norms only in
situations which allow for alternative kinds of behaviour, involving the need to
select among these, with the additional condition that selection be non-random.
In as much as a norm is active and effective, one can therefore distinguish
regularity of behaviour in recurrent situations of the same type, which is the
clearest manifestation of the 'order and predictability' Davis (1994: 97) regards
as characteristic of social creativity (see paragraph 3.1).

Needless to say, whatever regularities are observed, they themselves are not
the norms. They are only external evidence of the latter's activity, from which the
norms themselves (that is, the 'instructions' which yielded those regularities) are
still to be extracted; whether by scholars wishing to get to the bottom of a
norm-governed behaviour or by persons wishing to be accepted in the group and
hence needing to undergo socialisation (paragraph 3.7).

There is an interesting reversal of direction here: whereas in actual practice, it
is subjugation to norms that breeds norm-governed behaviour which then results
in regularities of surface realisations, the search for norms within any scholarly
programme must proceed the other way around. Thus, it is regularities in the
observable results of a particular kind of behaviour, assumed to have been

A Handful of Paragraphs on Translation and 'Norms'

[pg.] 17

governed by norms, which are first noted. Only then does one go on to extract
the norms themselves, on the (not all that straightforward) assumption that
observed regularities testify to recurrent underlying motives, and in a direct
manner, at that. For the researcher norms thus emerge as explanatory hypotheses
(of observed [results of] behaviour) rather than entities in their own right.

3.7 Norms and sanctions

In an established group, norms are basically acquired by the individual -- a
newcomer to the group on whatever grounds -- in the process of his/her
socialisation (paragraph 3.6). Very often, these norms -- or even the basic
agreements and conventions -- go on being negotiated throughout one's entire
life in the group, for instance, when members struggle to establish their own
positions within the group (or vis-à-vis its other members). Moreover, certain
individuals may be more instrumental than others in effecting changes in the
norms, depending on the status and position they have acquired in the group.

Be that as it may, unlike the weaker, more obscure conventions (paragraph
3.3), the notion of norms always implies sanctions; actual or at least potential,
whether negative (to those who violate them) or positive (to those who abide by
them). Within the group, norms also serve as a yardstick according to which
instances of behaviour and/or their results are evaluated, the second, comple-
mentary role any kind of norms is desi~gned to fulfil.

3.8 The graded and relative nature of norms

The instruction-like constraints of the norm type are far from monolithic: not
only are some of them more binding than others, at any single point in time, but
their validity and relative strength are bound to change over time.

Firstly, in terms of their potency, constraints on behaviour can be described
along a scale anchored between two extremes: general, relatively absolute rules
on the one hand, and pure idiosyncrasies on the other. The vast ground between
the two extremes is occupied by norms, which, in turn, form a graded continuum:
some are more rule-like, others -- almost idiosyncratic. In fact, we -- members
of a group as well as observers from without, including researchers -- can
recognise a mode of behaviour or its results as being idiosyncratic (or inevitable,
for that matter) only against the backdrop of our acquaintance with the middle
ground and its infernal gradation. Nor is the centrality of norms metaphoric only,
in terms of their relative position along a posited continuum. In a very strong
sense, the other two types of constraints are mere variations of norms, and not
independent entities. Consequently, they may easily -- and justifiably -- be
redefined in their terms: rules as '[more] objective', idiosyncrasies as '[more]
subjective [or: less inter-subjective]' norms.

Secondly, the borderlines between the various types of constraints are diffuse.
Each of the concepts, including the grading itself, is relative too, depending on
the point of view from which they are regarded, or the context into which they
are entered. Thus, what is just a favoured mode of behaviour within a large
and/or heterogeneous group may well be assigned much more binding force
within a particular subgroup thereof, which is likely to be more homogeneous
too (e.g. translators among text-producers, translators of literature among

[pg.] 18

Current Issues in Language and Society

translators, translators of poetry among translators of literature, translators
active in a systemic centre vs. translators who operate on a periphery, etc.). A
similar kind of relativity can be discerned in terms of types of activity, forming
either parts of each other (e.g. interpreting, or legal translation, within translation
at large) or just sharing adjacent territories (e.g. translation criticism vs. actual
translation [paragraph 4.9]). Thus, even if it is one and the same person who
engages in more than one activity, and/or belongs to more than one (sub) group,
s/he may well abide by different norms, and manifest different kinds of
behaviour, in each one of his/her roles and social contexts. The ability to
manoeuvre between alternative sets of norms is of course an important aspect of
social life, and its acquisition is an important component of socialisation.

Thirdly, along the temporal axis, each type of constraint may and often does
move into its neighbouring domain(s) through processes of rise and decline.
Thus, under certain circumstances (which would have to be specified), mere
whims may catch on and become more and more binding, and norms can gain
so much validity that, for all practical purposes, they become as strong as rules.
This may also happen the other way around, of course: what used to be binding
may lose much of its force, what used to be common may become rare, what was
once common to many may become idiosyncratic, on occasion even bizarre.
Needless to say, what was taken up in basically synchronic terms under the
second point can also be projected on the diachronic axis, which compounds the
possibilities as well as the difficulties inherent in the scholarly hunt for norms.

3.9 Norms and power relations

As already indicated (paragraph 3.7), shifts of validity and potency have a lot
to do with changes of status, and hence with power relations; whether these occur
within the group itself or whether power is imposed on it from without (a claim
which -- when stripped of its ideological overtones -- is far from an innovation
of postmodernist, feminist, post-colonialist and suchlike approaches to society
and culture). Whatever these shifts, they can always be accounted for in
connection with the notion of norm, especially since, in as much as the process
goes on and social agreements are re-negotiated, the constraints are likely to cross
its realm (paragraph 3.8), i.e. actually become norms, at least for the time being.

Having covered some general ground, let us move on to tying the notions of
social agreements, conventions and (especially) norms to the particular kind of
behaviour which translation appears to be.

4 Norms of Translation

4.1 Does translation carry the notion of 'norm'?

It has often been claimed that every art of translation involves a unique
encounter of an individual with a text within a specific communication situation.
Does such a view of translation carry the notion of 'norm' at all?
'The scope of sociability covers all our activities', says Davis (1994: 97); and
translation is certainly an activity which is of sociocultural relevance.

We try to do all these things in a conventional way, and when we agree that
we have options we try to create conventional ways of deciding among

A Handful of Paragraphs on Translation and 'Norms'

[pg.] 19

them. And you should note that creative activity is continuous: [conven-
tionalised kinds of behaviour] would cease to happen if we did not, so to
speak, renew the understanding which makes them, each time [...]. (Davis,
1994: 97)

In fact, it is not difficult to see why translation should lend itself to treatment
in terms of sociability, and quite easily so. After all, it is basically performed
within a sociocultural context, more often than not for the consumption of
persons other than the producers of the translated texts themselves, who may be
said to form some kind of a group together (paragraph 4.4). While there does
exist the notion of socially-insignificant translation (i.e. individuals translating
for themselves, so to speak; e.g. Harris & Sherwood, 1978), its practice is surely
negligible. Moreover -- and more importantly, in this context -- most socially-
redundant instances of translation can be expected to simulate socially-relevant
ones anyway, wittingly or unwittingly. Consequently, norms are bound to affect
them too, and the same norms, at that (which is one important way how potential
sanctions may be said to be taken into account [paragraph 3.7]).

4.2 Acts of translation vs. translation events

But is translation not a cognitive process? Does it take place anywhere else
except in an individual's brain? And if this is the case, what could the explanatory
power of sociocultural notions such as conventions and norms possibly be?
Should acts of translation not be accounted for in purely mental terms? Should
we not all turn to the cognitive sciences, if what we are interested in finding out
is what translation activities consist of and how they proceed in 'real life'?

True enough. All translation decisions are made in an individual's brain. At
the same time, positing an incongruity, let alone contradiction, between the
cognitive and the sociocultural seems a gross exaggeration, especially in the
context of translation; an incongruity which diminishes to the point of losing its
pointed tip as soon as a distinction is drawn between the act of translation, which
is indeed cognitive, and the context of situation where the person performing the
act, and hence the art itself, are embedded, which has sometimes (e.g. Toury,
1995: 249ff.) been called the translation event.

Needless to say, no translation event can be said to have taken place unless an
art of translation was indeed performed. On the other hand, at least in
socially-relevant instances of translation, including simulated ones (paragraph
4.1), all cognitive processes occur within contexts which constitute events. This
much I believe should be taken for granted, and it should be justification enough
for approaching the overall event in sociocultural terms.

One thing I would not venture to do here is tackle the intriguing question of
how, and to what extent, the environment affects the workings of the brain, or
how the cognitive is influenced by the sociocultural, even though this would
surely make an invaluable contribution to our understanding of translation (see,
in this connection, recent attempts to use the notion of 'meme' in Translation
Studies, especially to account for changes in the concept of translation itself and
the way they travel; most notably in Vermeer, 1997 and Chesterman, 1997.) We
will return to the possible influence of the environment on translation perform-
ance in the more specific context of socialisation; in this case -- the emergence of

[pg.] 20

Current Issues in Language and Society

an individual translator within an established sociocultural setting (paragraph

4.3 'All is predestined -- but freedom of choice is granted'

But is the concept of norm, especially in its application to translators, and hence
its association with translational behaviour, not too rigid, as many seem to
maintain? Is it really the case that acceptance of the idea that translation events-
are basically norm-governed entails the denial of free choice during an act of
translation which is embedded in it? -- Not at all! To borrow a concept from
traditional Jewish thinking: 'All is predestined but freedom of choice is granted'
(Avot (The Sayings of the Fathers ], 3: 15).

To put it bluntly, it is always the translator herself or himself, as an
autonomous individual, who decides how to behave, be that decision fully
conscious or not. Whatever the degree of awareness, it is s/he who will also have
to bear the consequences. Remember the notion of sanctions (paragraph 3.7) ? At
the same time, it is dear that, even though there is always the possibility that one
would be willing to take the risks which unconventional, non-normative
decisions entail, under normal conditions, a translator would tend to avoid
negative sanctions on 'improper' behaviour as much as obtain the rewards which
go with a 'proper' one. Needless to say, it would make an interesting project to
study the [negative and positive] sanctions that may be associated with
translational behaviour and their (possible and actual) effects on instances of
performance within defined sociocultural settings (see the way Daniel Simeoni
has recently related the concept of 'norm' and Bourdieu's notion of 'habitus' in
the specific context of translation, Simeoni, 1998).

In fact, translators have even been known to art differently, or at least to
produce different surface realisations of the category 'translation' (i.e. differently
looking utterances), when working for different commissioners, e.g. in order to
be given more work by the same commissioners, or at least to escape the need to
have their products edited by others, which many translators abhor. To be sure,
freedom of choice is exerted not only when one chooses to behave in a way which
does not concur with the prevailing norms. It is also exercised when one seems
simply to reaffirm one's previous commitment to these. After all, in principle,
there is always an alternative, otherwise there would be no need for norms in the
first place (paragraph 3.6).

4.4 What group is it where agreements are negotiated?

If agreements and conventions are constantly being negotiated (paragraph
3.2), and if norms are one of their outcomes and modes of implementation in
actual behaviour (paragraph 3.3), it would only be proper to enquire as to where
those negotiations take place, in the case of translation; in other words, what
constitutes 'the group' in question. For instance,

* How homogeneous (or heterogeneous) should that group be taken to be?
* Would it always consist of members of the same categories?
* More specifically, would it include acting translators only (which would
yield a very limited group indeed), or persons playing other, adjacent roles
as well; whether in the production of translations itself (e.g. editors [of

A Handful of Paragraphs on Translation and 'Norms'

[pg.] 21

translations, or even texts in general]; teachers, especially of translation;
translation critics; censors; publishers) or around it?
* And what about (average or specific) consumers of translated utterances:
would they be taken to form part of the group too? And, if so, would it not
mean going way too far with the notion of 'group'?
* In a heterogeneous group, how powerful are the translators themselves,
with respect to the creation, negotiation, maintenance and change of
translational norms? Do they occupy the centre or a peripheral position?
* And what about individuals who play several roles alternately? Is it all that
certain that they would art according to exactly the same norms while
assuming their different positions?

This is an intriguing domain about which translation scholars seem to know
precious little, beyond a small number of accounts of isolated individual cases.
There seem to be many alternative patterns here. Thus, there may be larger and
smaller groups involved in the negotiations, more or less closely-knit or diffuse,
more or less homogeneous, more rigid in their (personal or sectorial) composi-
tion, etc. It is not totally unjustified to assume that these differences would
manifest themselves as significant, in their implications for translation behaviour
and the norms governing it. However, this is about all I would say, at this point
(see again Simeoni, 1998 ).

4.5 The 'value' behind translation

Translation is a kind of activity which inevitably confronts different languages
and cultural traditions, and hence different conventions and norms on each
pertinent level. Thus, the value behind it, or the basic tools in a translator's 'tool
kit', for those who refuse to accept that values behave as causal elements of
culture (paragraph 3.4), may be described as consisting of two major elements;
namely, producing a text in a certain (so-called 'target') language,

(1) which is designed to occupy a certain position, or fill in a certain slot, in the
culture that uses that language while, at the same time,
(2) constituting a representation in that language/culture of another, preexist-
ing text in some other language, belonging to some other culture and
occupying a definable position within it.

It is clear that these two types of requirement derive from two sources which
-- even though the actual discrepancy between them may vary greatly -- should
be regarded as different in principle. Often they are incompatible in practice too,
so that any attempt to abide by the one requires a price in terms of the other,
which breeds an inherent need for compromise.

4.6 Norms and efficacy

The inevitable compromise between the constraints drawing on the two
different sources, while always realised by an individual, is strongly affected by
sociocultural factors which determine its appropriateness; the behaviour by the
enveloping circumstances, the art by the event. Among other things, this can be
seen as a strong factor of efficacy.

Thus, were it not for the regulative capacity of norms, the tensions between

[pg.] 22

Current Issues in Language and Society

the two sources of constraints, and hence between adequacy and acceptability
(as a translation or as a target-language text) would have to be resolved on an
entirely ad hoc basis, and with no clear yardstick to go by. Indiscriminate, totally
free variation might have been the result, which would have made it next to
impossible to locate an art of translation and/or its results within their social
context and assign them any cultural relevance. Everything may well have been
seen as equal to everything else, which is most certainly never the case; not even -
in the most permissive of societies. Not even if the ideology appears to be there.

4.7 Regularities of behaviour again

In fact, as any cursory look will ascertain, translation as practised within a
particular culture, or a certain sector thereof, tends to manifest quite a number of
regularities, in terms of both translational adequacy and acceptability (as well as
their preferred blends), a fact which we have already taken as strong evidence of
the potency of norms (paragraph 3.8). On the other hand, these regularities may
well differ from the ones exhibited by another culture, cultural sector, or even the
same culture in another phase of its evolution (which amount to the same thing,
theoretically speaking).

One consequence of the existence of such regularities and their acknow-
ledgment is that, even if they are unable to account for them, people-in-
the-culture can at least tell when a translator has failed to adhere to sanctioned
practices. For instance, they may not be able to say that a certain phenomenon in
a translated text reflects interference from the source text/language, but they will
at least have a hunch as to what they are expected to~ feel about it, within the
preferences of their culture. Different cultures have been known to have had
different thresholds of tolerance of interference. Some of them even preferred to
have them in translated texts; e.g. in the translation of 'Works of Wisdom' (in
contradistinction to the translation of 'Works of Beauty') into Hebrew in the
Middle Ages (see Toury, 1998).

4.8 How regular would 'regularities' need to be?

'Regularities' thus turn out to be a key notion in descriptive studies into
translational behaviour and its results as well. In fact, the establishment of
recurrent patterns is the most basic activity in the pre-explanatory phases of a
study, the phases where data are collected and analysed and discoveries are being
made. Also, it is first and foremost discerned regularities, rather than any of the
individual phenomena as such, which would then be explained on the assump-
tion that the behaviour which yielded them was indeed norm-governed
(paragraph 3.6), all the more so as not every observed phenomenon will be
subsumable under one of the emerging recurrent patterns in the first place. The
beauty of human behaviour, whether under cognitive or sociocultural observa-
tion, is that there is no 100 per cent regularity, not even in the behaviour of one
person while translating one text, which concurs with the graded and relative
nature of the notion of norms (paragraph 3.8). Complete absence of any
regularities should also be regarded as marginal: if one looks hard enough (or
extends one's corpus enough), reflections of any possible mode of behaviour are
sure to be found.

A Handful of Paragraphs on Translation and 'Norms'

[pg.] 23

What may seem more frustrating than failure to come up with any absolute
findings (which both 'never' and 'always' imply) is that, very often, regularities
will first manifest themselves in rather low percentages. Consequently, it will not
be all that clear just how much significance should be assigned to each observed
regularity. The main reason is that it is almost as hard eta establish sampling rules
for translational behaviour, or even its textual-linguistic results, as it is to take
'everything' into account. Justifying the status of a body of material as a 'sample'
in terms of Translation Studies is even harder.

In actual fact, what a researcher often starts out with is a rather arbitrary set
rather than a proper corpus; a group of texts, or a number of lower-level
phenomena, which may be both accidental, from a translational point of view,
and highly heterogeneous (i.e. devoid of clear regular patterns). The way to go
from here is to try and break the initial set into sub-groups on the basis of one
feature (variable) or another which will have emerged as significant (for that set)
during the study itself. This procedure is bound to yield a substantial increase of
homogeneity, reducing each sub-group's accidentality and gradually rendering
it representative in terms of that particular variable; in other words, a proper
corpus. Within such sub-groups, regularities are bound to increase, often
considerably. If found to be too small now, any subgroup-turned-corpus could
then be expanded; this time on the basis of the defining feature itself, and hence
in a much more justified (and justifiable) fashion.

4.9 Are translational norms translation-specific?

Due to their contending sources (paragraph 4.5), there is no way that the norms
governing translation in their totality (that is, the overall 'normative model' a
translation event is subject to) will be identical to the ones operating in any other
field, be it even a closely-related one. One may of course expert correlations,
including partial overlaps, but never full identity. Norms can also be imported
from one type of behaviour to another (always with some [necessary] modifica-
tions), but the value of each one of these norms is likely to be different due to its
different systemic position. The same holds for norms imported from a different
group engaging in the same kind of activity, within the same culture and society
or in different ones.

Let us look at three types of activity which are closely related to translation:
communication in non-translated utterances, translation assessment and trans-
lator training.

Type one: Communication in translated vs. non-translated utterances. Here, partial
overlap is to be expected, as a translation is always an intended utterance in the
target language and culture: one aspect (or phase) of any art of translation
involves formulation in that language, and the norms governing this activity may
of course be more or less similar to the norms governing the composition of a
non-translated utterance, and more or less different from them. At the same time,
since translation is not reducible to that aspect/phase alone, only partial overlap
can be expected.

Overlap between the norms governing translation and non-translation can be
expected to grow in direct proportion to the centrality of target-language
'normality' in translation, which is of course a norm-governed idea. 'Accept-

[pg.] 24

Current Issues in Language and Society

ability as a translation' may thus become a variety of 'acceptability' in general,
which would normally imply reduced interest in the principles governing the
source text (or the internal 'web of relationships' which constitutes it) and their

Conversely, imitations of textual-linguistic behaviour in another culture/lan-
guage (or of certain types of translation from it) may be attempted in
non-translational communication as well. The resulting texts may thus bear close
resemblance to translations without there ever having been an identifiable source
text. This option has often been selected by the creators of so-called 'fictitious
translations' (e.g. Toury, 1995: 40-52), precisely in order to convey the impression
that those texts had in fact been translated, i.e. lead the people-in-the-culture
astray on the basis of what they have come to associate with 'genuine'

Whether the one extreme or the other (and both of them are extremes!), the
likely result is a blurring of the borderline between translations and non-transla-
tions: all texts would tend to look alike, and whatever differences there may be,
it won't be easy to attribute them to the translation/non-translation opposition.
On occasions, such an opposition may be found out to have been completely
non-functional in the culture in question: even if it is retained on the level of the
acts whereby texts are generated, translations may still be presented -- and
accepted -- as originals, and originals as translations, without this having any
cultural repercussions.

Such seems to have been the case in the early Enlightenment period in Hebrew
literature (Toury, 1995: 131ff.). People-in-the-culture, producers and consumers
alike, did not really rare which texts were based on foreign ones in a one-to-one
manner and which texts were based on them in a one-to-many ratio, or even just
embodying the principles of the models underlying particular text-types.
Today's scholars face a great many difficulties in ascertaining which one is which,
but, in an important sense, they may be trying to salve riddles of very little
historical significance, if that is what they are doing.

Type two: Translation vs. translation assessment. These two activities differ in a
different sense: with respect to the translational product, the one activity is
prospective, the other one is retrospective. Even if bath result in textual entities, their
end-products are of a different order too: translations are the result of a direct
application of translational norms whereas assessments employ first and
foremost norms of evaluation and of evaluation-presentation, including the
norms governing the composition of evaluative texts. As regards translational
norms, evaluators just react to them and their results. Sometimes they may try to
extract them from the results of translational behaviour, to a certain extent even
verbalise them. What they never do is implement the norms, unless they wish to
offer an alternative translation (which may be a strategy of critics, even teachers,
whereby they change for some time their role, and hence the kind of activity they
are engaging in).

Basically, translation and evaluation are two different activities, then, whose
governing principles can simply never be 'the same'. What they can do is reflect
the same overall attitude towards translation, each in its own domain. In this
sense, translators and evaluators may belong to the same group (paragraph 4.4).

A Handful of Paragraphs on Translation and 'Norms'

[pg.] 25

However, even this is not a theoretical must, nor does it always occur in practice.
Thus, critics and translators may, and often do have different values, e.g. they may
favour different blends of acceptability and adequacy; on occasion, even if
translation and criticism are performed by one and the same person.

To give an example: whether (or under what circumstances) a translator would
feel obliged to read the source text in its entirety before s/he embarks on its
translation is a basic strategic decision which may rely more or less on social
factors, that is, be norm-governed; be it directly (e.g. there exists an 'instruction'
to do precisely that) or indirectly (e.g. through a marked dominance of an initial
norm of translation adequacy). Now, whereas every translator who has finished
translating will have been through the source text at least once, many translation
evaluators, including critics writing on literary translations, or members of
committees awarding translation prizes, may never feel an urge to even peep into
the original, let alone read it in its entirety. What is most significant here is not
simply that this happens, but that a societal group may accept it, sometimes even
prefer it that way.

Type three: Translation vs. translator training. This may well be the trickiest
comparison. One would think that persons who have been entrusted by society
with the training of translators at an accelerated pare, would see their task as
imparting modes of behaviour to the non-initiated the way they are normally
practised, thus preparing them for acceptance in and by the relevant group
(paragraph 4.4). However, this is often not the case. What many students of
translation are actually being offered draws on an admixture of concepts
borrowed from sources deemed more 'respectable' than the behaviour of real
translators under normal sociocultural conditions; mainly disciplines such as
linguistics, text-linguistics or pragmatics. These concepts are supplemented by
intuitions, sometimes very good ones, but all too often seasoned with a speck of
wishful thinking. In the most extreme cases, the claim is even made (at least
implicitly) that there are things that simply should be (or else should never be)
done; by virtue of what translation allegedly 'is', and not by virtue of a sheer
convention; in fact, not seldom in contrast to existing conventions and the
agreements which underlie them.

There is normally an ideology behind such attitudes, and ideologies tend to
involve a manipulation of existing normative patterns. Thus, many teachers of
translation see it as their task to effect changes in the world at large, wishing, as
it were, to take active part in the process of (re-)negotiation which is constantly
going on (paragraph 3.2). They would of course claim that the prevailing
situation is badly in need of improvement, but this would not affect the basic
claim that what they are trying to do is change a state of affairs, and one which
others, including the group of practising translators, may well regard as perfectly
satisfactory. They do what they do from a position of almost absolute power
(vis-à-vis their students), power which was granted to them by the institutions in
which they work; sometimes, though not always, and certainly not necessarily,
on the basis of their own recognition by society as translators (see, in this
connection, Chesterman's 'professional norms', Chesterman, 1993). However,
the edge teachers have over their students does not necessarily imply similar

[pg.] 26

Current Issues in Language and Society

position and power within society at large; not even in the 'translatorial' group,
whoever it may consist of.

It is possible to say that training institutes often behave like closed groups,
having conventions and norms of their own. They are trying to impart these
norms to newcomers to this closed group and through them and their future
translational activity -- to society as a whole. Unfortunately, transition from such
a group into the 'world' may not always be all that smooth. In extreme cases it
may involve real pain and frustration. Thus, it has not infrequently been the case
that the graduates of a translation programme had to undergo a process of
forgetting a great deal of what they had been taught and adjusting, at least in
part, to prevalent norms of sociocultural appropriateness; very often the very
same norms their teachers wished to see changed.

4.10 But are all translations 'good'?

Are teachers of translation (or critics, for that matter) all that wrong? Has
everything become so relative that there is no such thing as a bad translation any
more? Of course there is, even though there is definitely nothing objective or
absolute about that notion either. Rather, notions of what would constitute a bad
translation (or a good one, for that matter) are as changing as the notion of
translation itself. In fact, judged by our (irrelevant) norms, even the ones we apply
to the case of the translation of religious texts, the King James Version of the Bible
is surely not a very good translation. From today's point of view, its centrality
can only be explained in historical terms. Acknowledging its 'inherent' qualifies
as bath an English text and a translation into English would require the adoption
of another attitude, associated with a completely different set of norms, those that
were at play at a different point of time and hence in another culture.

The basic thing one must be ready to accept is that bad translations are first
and foremost translations, not something completely different. Consequently,
whether an item which would be conceived of as a translation is 'good' or 'bad'
will be determined by an extension (or further specification) of the normative
model pertinent to the culture where it came into being (or the appropriate
section within it). It is not that members of a societal group cannot arrive at a
valid conclusion in an intuitive way; it is that, if and when required to account
for their attitude, they will have to draw on that set of norms -- or else be unable
to justify their intuitive verdict.

Thus, any attempt to impart the way 'good' translations are (to be) done, e.g.
by teachers of translation (paragraph 4.9), may backfire; namely, when society
refuses to accept that those are indeed good. Significantly enough, it is often the
case that even teachers in one and the same institute, or critics within one culture,
do not assess a translation the same way. Differences of assessment, again, may
look idiosyncratic, a matter of personal taste and temperament, and to a certain
extent this is precisely what they are. However, to an even greater extent they are
a result, or a reflection of the affiliation of different persons to different
(sub)groups, and to that extent they are a function of norms again. (For the
possibility of having different, even competing norms within one and the same
group see paragraphs 4.12 and 4.13.)

A Handful of Paragraphs on Translation and 'Norms'

[pg.] 27

4.11 How does an individual acquire translational norms?

In modern times, many translators (but still a minority) are indeed being
trained, even conditioned, in professional or academic training institutes. Me
have just touched upon their possible fate (paragraph 4.9). Others, probably the
vast majority, pick up the conventions and norms pertinent to their job through
a process of initiation within the culture itself, a specific mode of socialisation
(paragraphs 3.6-3.7). In view of the lack of any real longitudinal studies into the
making of translators outside of the schooling system, the only way to sketch this
process is speculative; namely, on the basis of what we know about socialisation
in general, with the addition of some translation-specific considerations. The
present writer also made use of his insights as to how he himself became a
translator back in the 1960s; namely, before he ever did any Translation Studies.

Firstly, if the internalisation of norms is really that important an aspect of
translational behaviour, then the acquisition of that knowledge, and of ways of
coping with it in real-life situations, should count as a major aspect of
socialisation in relation to translating. My assumption here is that, being a mode
of communication, translating is likely to involve environmental feedback, which
may come from any other party to the communication event. This feedback is
normative in its very essence: it concerns the well-formedness of a translation,
not just as an utterance in the receptor language and culture, but first and
foremost as an assumed translation (Toury, 1995: 31-35), that is, a realisation in
the culture and language in question of the mode of text-production translating
is taken to be. At least by implication, the norms embodied in that feedback also
apply to the (minimal, optimal, necessary, etc.) relationships between assumedly
translated utterances and their assumed sources, especially in terms of whatever
should have remained invariant. By extension, they also determine the appropri-
ateness of the strategies used to derive a translational output from a given input
utterance under those conditions of invariance, even though there can be no
one-to-one relation between a procedure and the results of its application.

Secondly, it is the all-pervasiveness of sanctions (paragraph 3.7) which lends
such normative feedback its influence on a translator's behaviour. Under normal
conditions, one would wish to avoid negative sanctions on 'improper' behaviour
as much as obtain the rewards which go with a 'proper' one (paragraph 4.3). This
aspiration holds especially for the novice, who -- due to lack of sufficient
experience -- is likely to feel insecure as to what translating is all about, according
to the conception of the group in and for which s/he will be operating, and who,
on the other hand, may be looking for recognition by that group in his/her
capacity as a (socially-relevant) translator. It is precisely this view which
gradually crystallises for her/him in a process of initiation. It may, of course,
prove irrelevant again, if and when that person moves to another group,
especially if the new group forms part of a completely different culture. Under
such circumstances, another process of socialisation may be required.

Thirdly, in the initial stages of one's development as a translator, the feedback
directed at him/her is exclusively external: overt responses to one's translational
products, final or interim. A novice simply has no means of assessing the
appropriateness of various options and/or of the alternative strategies that may
yield them. Little by little, however, translators may start taking potential

[pg.] 28

Cunent Issues in Language and Society

responses into account too. They thus develop an internal kind of monitoring
mechanism, which can operate on the (interim) product as well as on the art of
translation as such.

Fourthly, as socialisation in relation to translating goes on, parts of the
normatively-motivated feedback are probably assimilated by the translators as
they gather more and more experience, modifying their basic (i.e. innate)
competence and gradually becoming part of it. Many decisions will now be made
more or less automatically. It may also be hypothesised that, to the extent that a
norm has been internalised and made part of a modified competence, it will be
applied to instances of more spontaneous translation too, namely in situations
where no sanctions are likely to be imposed. It is in this sense that socially-insig-
nificant instances of translation may be said to simulate socially-significant ones
(paragraph 4.1). Some translators may then go on to take active part in the
re-negotiations concerning translational conventions (paragraph 3.2) which will
sometimes result in a change of norms.

4.12 Alternative norms within a group

One thing which makes translational decision-making less demanding than it
may have sounded so far in terms of the risks taken, even though probably more
complex in terms of its underlying mechanisms, is the fact that, at every point in
the life of a societal group, especially a comprehensive and/or heterogeneous
one, there tends to be more than one norm with respect to any behavioural
dimension. Consequently, the need to choose between alternative modes of
behaviour tends to be built into the very system, so that socialisation as concerns
translating often includes acquisition of the ability to manoeuvre efficiently
between the alternatives (paragraph 4.6).

Multiplicity of norms does not amount to no norms at all, much less imply
anarchy. For it is normally not the case that all existing norms are of an equal
status, so that choice between them would be totally free, or devoid of any
implications for the assessment of the person's behaviour and/or his/her
position within society. Manoeuvring between alternative modes of behaviour
thus turns out to be just another norm-governed activity, necessarily involving
risks of its own.

4.13 Competing norms

Norms operating within one and the same group are not merely different from
each other. Quite often they are competing too. After all, in the dynamic structure
of a living society, there is always a struggle for domination, as a result of which
norms may change their position vis-à-vis a certain centre of gravity, the more so
as the centre itself may be undergoing shifts.

What complicates matters even more is the fact that each group-within-a-
group (and all groups tend to be hierarchically organised) may have its own
structure of centre vs. periphery, entailing an internal struggle for domination of
its own, in addition to (and sometimes as part of) its participation in the overall
struggle. Consequently, one has to be as clear as possible as to whether one is
talking about changes of (sub)systems or changes within one of them.

Firstly, there is variation within a culture. Whether within one (sub)system or

A Handful of Paragraphs on Translation and 'Norms'

[pg.] 29

between the various (sub)systems regarded as building up one higher-order
entity, it is not rare to find side by side three types of competing norms, each
having its own followers and a position of its own: the ones that dominate the
centre, and hence direct translational behaviour of what is recognised as the
mainstream, alongside the remnants of previous sets of norms and the rudiments
of what may eventually become new ones, hovering in the periphery (and/or
near the centre of lower-order (sub)systems). This is why it is possible to speak
-- and not derogatorily either -- of being 'trendy', 'old-fashioned' or 'progres-
sive' in translation as it is in any other behavioural domain.

Secondly, there are changes over time. One's status as a translator, in terms of
the norms one adheres to, may of course be temporary: many translators fail to
adjust to the changing requirements, or do so to an extent which is deemed
insufficient. Thus, as changes of norms occur, formerly progressive translators
may soon find themselves just trendy, on occasion even downright passé.

At the same time, regarding this process as involving a mere alternation of
generations can be misleading, especially if generations are equated with age
groups. While there often are correlations between one's position along the
'dated'-'mainstream'-'avant-garde' axis and one's age, these cannot, and should
not be taken as inevitable, much less as a starting point for the study of 'norms
in action'. As already maintained, there is nothing deterministic here.

In fact, research shows that it is often people who are in the early phases of
their initiation as translators, whether young or not so young of age, who behave
in the most epigonic way. Insecure as most of them understandably are, they like
to play it safe and tend to perform according to dated, but still valid norms. One
way to explain this is to realise that a beginner's deviant behaviour would more
readily be regarded by society as 'erroneous' rather than 'innovative'. while both
may be applied to the same mode of behaviour (or its products), the different
values assigned to them make all the difference in the world!

Such a conservative tendency is further enhanced if would-be translators
receive reinforcement from socialisation agents, especially powerful ones,
holding to dated norms themselves. No wonder that revolutions -- i.e.
large-scale changes of paradigm -- have often been made by experienced
translators who had, moreover, attained considerable prestige by behaving
'appropriately', i.e. according to mainstream norms. After they internalised those
norms, and having attained more than mere recognition by society, they can
afford to start deviating from them and get away with it.

4.14 Constraints, strategies and norms

There is another pair of notions which deserves a lot more attention than it has
received so far, namely that of 'strategy' and 'norm'. Let us regard as a strategy
any set of moves utilised in trying to solve a perceived problem; perceived by the
one performing the act, that is.

Intuitively, there seems to be some connection between strategies, on the one
hand, and norms on the other. However, the nature of that connection has not
been clarified; probably mainly due to the fact that those who focused on
translation strategies (e.g. Wolfgang Lörscher, 1991) have normally considered
mere acts of translation rather than translation events (paragraph 4.2) whereas most

[pg.] 30

Current Issues in Language and Society

socioculturally-oriented scholars have not followed the progression from the
event to the art. I believe it is about time that we had both ends meet, if only for
the purpose of supplying better, more comprehensive and more flexible
explanations of the translational behaviour of individuals within a societal
context (and see Simeoni, 1998 once again). Such explanations are also bound to
assign a third notion, that of constraints, its proper place in the account of
translation rather than its mere 'opening conditions'.

5 By Way of Conclusion: A Story of Three 'Killers'

Let me conclude with one of my favourite cases, which may be taken as a
nut-shell exemplification of many of the points made throughout this document.

In the last few decades, three different Hebrew translations of Hemingway's
famous short story 'The Killers' were published, at almost identical intervals, and
not very long ones, at that: the first translation (A) was published in 1955, the
second one (B) in 1973 and the third and last one (C) in 1988. Linguistically, each
one of the three textual entities is of course different, which is all but surprising;
the more so as every translator seems to have been aware of the earlier version(s).
In fact, some of the decisions made by later translators could be taken as
indications of so-called 'polemical translation'; see Popovic, 1976: 21.)

What is most interesting, however, and not all that evident, is that when asked
to put the three translations in their correct chronological order, everybody --
from complete newcomers to thinking about translation to experienced transla-
tors, teachers of translation and translation scholars -- came up with precisely
the same order. Moreover, with very few cases of local disagreement, when asked
to justify their ordering, they all based it on the same series of features; basically
an assortment of semantic, grammatical, syntactic, pragmatic and stylistic
markers, as well as translation relationships, which the subjects seem to have
associated with 'typical behaviour' of literary translators into Hebrew at the three
different points in time (or at least of its gradual change along time).

Now, the significant thing is that, in spite of a 100 per cent agreement between
dozens of subjects who have undergone this pseudo-experiment, they were all
wrong: the order they came up with -- which was based on their intuitive-to-
learned ability to identify relevant markers and associate them with modes of
translation (and, yes, the norms which governed them) -- that order did not
conform to reality. Thus, the order they all gave was ACB instead of ABC.

When I disclosed the names of the translators (actually, four of them, because
the 1973 version (B) was prepared by two persons jointly), there were quite a
number of subjects who were able to correct their initial ordering. The names
acted as additional information for them, because they had cultural knowledge
as to who was more or less likely to count as 'dated', 'mainstream' or
'avant-garde' in their translational behaviour.

Finally, once all subjects learnt of their factual error, it was relatively easy to
explain to them how, historically rather than chronologically speaking (that is,
in terms of the appropriate norms and their position in society), they had not been
all that wrong, after all. Thus, only one of the three translations was appropriate
for the time at which it was produced and the expectations of its intended
consumers, namely, the first one (A). The other two versions were either ahead

A Handful of Paragraphs on Translation and 'Norms'

[pg.] 31

of their time (B) or somewhat obsolete (C); two kinds of deviation from
'mainstream norms' which were automatically 'corrected' by all readers.

The bottom line seems clear enough: not only are there norms associated with
translation, but people-in-the-culture know how to, and actually do activate
them; not only while producing translations themselves but while consuming
them as well. What is still unclear is whether production- and consumption-
norms are exactly the same, even in this individual case, i.e. which group it is that
generates and negotiates translational norms, but we have already presented this
as a moot point (paragraph 4.4).


1. I have decided to have the paper published in a format almost identical to that which
served as basis for the seminar, i.e. with only minor amendments. The reason behind
this decision is to let the readers see how I tried to stimulate the debate.


Chesterman, A. (1993) From 'is' to 'ought': Laws, norms and strategies in translation
studies. Target 5, 1-20.

Chesterman, A. (1997) Memes of Translation. Amsterdam and Philadelphia: Benjamins.

Davis, J. (1994) Social creativity. In C.M. Hann (ed.) When History Accelerates: Essays on
Rapid Social Change, Complexity and Creativity
(pp. 95-110). London and Atlantic
Highlands, NJ: The Athlone Press.

Even-Zohar, I. (1971) Introduction to a theory of literary translation. Unpublished PhD
thesis (in Hebrew), Tel Aviv University.

Even-Zohar, I. (1978) The position of translated literature within the literary polysystem.
In J.S Holmes et al. (eds) Literature and Translation: New Perspectives in Literary Studies
(pp. 117-27). Leuven: Acco. (Revised version in Even-Zohar, 1990: 45-51.)

Even-Zohar, I. (1990) Polysystem Studies. Tel Aviv: The Porter Institute for Poetics and
Semiotics, and Durham: Duke University Press (Poetics Today 11:1).

Geest, D. De (1992) The notion of 'system': Its theoretical importance and its
methodological implications for a functionalist translation theory. In H. Kittel (ed.)
Geschichte, System, Literarische Übersetzung/Histories, Systems, Literary Translations (pp.
32-45). Berlin: Erich Schmidt.

Gentzler, E. (1993) Contemporary Translation Theories. London and New York: Routledge.

Harris, B. (1990) Norms in interpretation. Target 2, 115-19.

Harris, B. and Sherwood, B. (1978) Translating as an innate skill. In D. Gerver and S.H.
Wallace (eds) Language, Interpretation and Communication (pp. 155-70). New York and
London: Plenum.

Hermans, T. (1991) Translational norms and correct translations. In K.M. van
Leuven-Zwart and T. Naaijkens (eds) Translation Studies: The State of the Art. Proceedings
of the First James S Holmes Symposium on Translation Studies
(pp. 155-69). Amsterdam
and Atlanta, GA: Rodopi.

Hermans, T. (1995) Toury's empiricism version one. The Translator 1, 215-23.

Holmes, J.S (1988) Translated! Papers on Literary Translation and Translation Studies.
Amsterdam: Rodopi.

Holmes, J.S, Lambert, J. and van den Broeck, R. (eds) (1978) Literature and Translation: New
Perspectives in Literary Studies
. Leuven: Acco.

Levy, J. (1969, 1963) Die literarische Übersetzung: Theorie einer Kunstgattung (Walter
Schamschula, trans.) Frankfurt am Main and Bonn: Athenäum.

Lörscher, W. (1991) Translation Performance, Translation Process, and Translation Strategies.
Tübingen: Narr.

Maschler, Y. (in press) Rotsè lishmoa kéta? ('wanna hear something weird/funny [lit. 'a
segment'] ?') : The discourse markers segmenting Israeli Hebrew talk-in-interaction. In
A. Jucker and Y. Ziv (eds) Discourse Markers. Amsterdam and Philadelphia: Benjamins.

[pg.] 32 Current Issues in Language and Society

Popovic, A. (1976) Dictionary for the Analysis of Literary Translation. Edmonton: The
University of Alberta, Department of Comparative Literature; Nitra: The Pedagogical
Faculty, Department of Literary Communication.

Shlesinger, M. (1989) Extending the theory of translation to interpretation: Norms as a
case in Point. Target 1, 111-15.

Simeoni, D. (1998) The pivotai status of the translator's habitus. Target 10, 1-39.

Swidler, A. (1986) Culture in action: Symbols and strategies. American Sociological Review
51, 273-86.

Toury, G. (1977) Translational Norms and Literary Translation into Hebrew, 1930-1945 (in
Hebrew). Tel Aviv: The Porter Institute for Poetics and Semiotics, Tel Aviv University.

Toury, G. (1978) The nature and role of norms in literary translation. In J.S Holmes et al.
(eds) Literature and Translation: New Perspectives in Literary Studies (pp. 83-100). Leuven:
Acco. (Revised version in Toury, 1995: 53-69.)

Toury, G. (1995) Descriptive Translation Studies and Beyond. Amsterdam and Philadelphia:

Toury, G. (1998) Hebrew tradition. In M. Baker (ed.) Encyclopedia of Translation Studies (pp.
439-48). London: Routledge.

Vermeer, H.J. (1997) Translation and the 'meme'. Target 9, 155-66.