Toward a Theory of Translating

Richards, I. A. 1953. Toward a Theory of Translating. In: Wright, Arthur F. (ed.), Studies in Chinese Thought. Chicago: The University of Chicago Press, 247-262.

The book was written hurriedly, in a whirl of lecturing on Ulysses and on The Possessed, durig a first teaching visit to Harvard. It was worked up from notes made between Tsing Hua and Yenching, under the guidance of diverse advisers, and written out with much of the feeling one has in trying to scribble down a dream before it fades away. (Richards 1953: 247)

That is almost like the way I'm writing about communication - in a hurried fashion, before the concept loses its clarity again and gets lost in a myriad of disconnected quotes and a vague cloud of ideas.

How can one compare a sentence in Englist poetry with one (however like it) in English prose? Or indeed any two sentences, or the same sentence, in different settings? What is synonymy? (Richards 1953: 249)

These are indeed different questions because the question of meaning in language and literature is open-ended. In terms of behaviour it is somewhat easier: when two forms or patterns of behaviour call forth the same interpretation then they are functionally the same. But by this token, much behaviour is actually meaningless and can stay that way unproblematically. But a sentence that doesn't mean anything is almost an oxymoron.

We may begin by adapting the conventional diagram of the communication engineer to our wider purposes. In translation we have two such diagrams to consider as a minimum. There will be (say) a Chinese communication for which we find ourselves in the role of Destination; and we assume thereupon the role of source for a communication in English. But since other communications in Chinese and other communications in English, having something in common with the present communication, come in to guide the encodings and decodings, the process becomes very complex. We have here indeed what may very well be the most complex type of event yyet produced in the evolution of the cosmos. (Richards 1953: 250)

This diagram is actually quite different from what I've seen elsewhere.

The comprehending of any utterance is guided by any number of partially similar situations in which partially similar utterances have occurred. More exactly, the comprehending is a function of the comparison fields from which it derives. Let the units of which these comparison fields consist to be utterances-within-situations - the utterance and its situation being partners in the network of transactions with other utterances in other situations which lends significance to the utterance. Partially similar utterances made within very different situations are likely to require different comprehendings, though language is, of course, our collective attempt to minimize these divergences of meaning.
A comprehending, accordingly, is an instance of a nexus established through past occurrences of partially similar utterances in partially similar situations - utterances and situations partially co-varying. The past utterances-within-situations need not have been consciously remarked or wittingly analized; still less need they be explicitly remembered when the comprehending occurs. Thus the word comparison in the technical term "comparison-field" may mislead. It is not necessary that the members of a comparison-field - widely diverse utterances-within-situations as they may be - should ever have been taken together in explicit analytic scrutiny and examined as to their likenesses and differences. The discriminations and connections (dividings and combinings) which arise in the development of meaning are, is some respects, as though this has been done. (Richards 1953: 251)

This sounds quite reasonable. Utterances and texts are indeed comprehended on the basis of previous utterances and texts.

What I have been sketching applies, for the translator, in the first place to the Decoding and Developing of the Chinese utterance. In the second place it applies to the Selecting and Encoding which (it is hoped) will produce an utterance in English acceptable as a translation from the Chinese. But, plainly enough, te co-varyings of utterances-within-situations for English are other than tehy are for Chinese. Any translator has acquired his Chinese and his English through "comparison-fields" which are different and systematically different in structure: different not only with respect to the ways in which utterances change with situations, but also with respect to those changes that are significant in utterances (e.g., phonemics) and with respect to those changes that are significant in situations (e.g., status recognition). The comparative linguist could, if he wished, illustrate this for the rest of his natural days. (Richards 1953: 252)

Richards takes note of the difference between studying significance in utterances (e.g. Jakobson) and social situations (e.g. Ruechs).

Let us turn our communication diagram 90 degrees now and look down it. Here is a cross-section of the activities to be found there, made at the points where what is prepared for transmission and what has been decoded and developed may be supposed - in a successful communication - to resemble one another most nearly. I have marked and numbered for labeling the seven divisions in my proposed schema.

I. Points to, selects. ...	I. Indicating
II. Says something about, sorts. ...	II. Characterizing
III. Comes alive to, wakes ut to, presents. ...	III. Realizing
IV. Cares about. ...	IV. Valuing
V. Would change or keep as it is. ...	V. Influencing
VI. Manages, directs, runs, administers itself. ...	VI. Controlling
VII. Seeks, pursues, tries, endeavors to be or to do. ...	VII. Purpose

Let us label these sorts of work which an utterance may be doing with two or more sets of names, academic or colloquial - on the assumption that communication will be made more probably if we use here a multiplicity of largely equivalent indications. I am numbering them for convenience of reference; but I do not want to suggest that there is any fixed temporal order, that fist we Select, then we Characterize, then Realize, then Value, then would Influence, then Organize and then Purpose. Nor is there any constant logical order. Let us keep these jobs as independent one of another as we can. (Richards 1953: 253)

I don't see how this schema is related to the diagram of communication - the 90 degree turn remains ambiguous. These semantic relations are valuable in themselves, though; and perhaps even comparable to Morris's types of discourse - in the notes of this article acknowledges his indebt to Morris's Signs, Language, and Behavior "for suggestions contributing to my schema".

The last paragraph illustrates - as must any attempt to write about the language we use or should use about language - the heavy duties we have put on quotation marks. I have suggested (How To Read a Page, pp. 68-70) that we should develop sets of specialized quotes, as a technical notation by which we could better keep track of the uses we are making of our words, and I have tried out the use of a few such quotation marks in that book and elsewhere. I am now more than ever persuaded of the usefulness of this device. It can serve us to distinguish many different uses we make of quotes. For example:

w...w	to show that it is a word - that word in general, Peirce's rtyper - which is being talked of. For example, wusew is a highly ambiguous word.
oc...oc	to show that occurrences of a word - Peirce's rtokenr - are being talked of. For example, I have been using ocusesoc above in various ways.
r...r	to show that some special use of the word or phrase is being referred to. the marks may be read as refer to and the implication may be that only by having that particular use of the word in that passage presents to us in lively attention (Realizing) can we distinguish it from other uses and avoid confusion.
t...t	to show that the word or phrase is to being used as a technical term anchored by a definition to some state of affairs or procedure - to an operational technique perhaps or to a set of performances.
q...q	to show how the word or phrase is to be comprehended is the question. It may be read as query; and we can develop this notation further by adding I-VII after the q to show wheret he focus of the question lies in my diagram. These q's should carry no derogatory suggestion; their work is to locate and orientate inquiry; they are servants of VI. Thus we might write qII connotationqIII,IV,V or qIII,IV,VconnotationqIII,IV,V to direct attentiot either to the logical or to the literary questions.
sw...sw	to show that we are considering what may be said with a certain word or phrase without decision as yet to what that is. This enables one to bring together meanings of words and phrases, for examination, without settling anything prematurely as to how they may be related. We need to bring these meanings together before we pick out those we may profitably compare. I have written elsewhere at length (Interpretation in Teaching, chaps. xv and xix; How to Read a Page, chap. x) on the troubles which the lack of such a warning mark may lead us into.
!...!	to show astonishment that people can write or talk so. Some will want to put this whole paper within such marks.

(Richards 1953: 255-256)

Wow. This is actually really cool. And with HTML it is easy to do: the specialized quotation marks go between ^and.

Sinclair's ^swgroupings^sw seems to be my ^swCharacterizing, Sorting^sw. We have, in English, what may seem an excess of analytic machineries to help us in distinguishing ^!its^! from ^!whats^!, that is, Indicating, I, from Charaterizing, II. Such are (in most uses): for I, ^wsubject, substance, entity, particular, thing, being, group, class^w; for II, ^wpredicate, attribute, property, quality, relation, character, essence, universal^w. A large methodological question which can seem to fall near the very heart of any endeavor to translate philosophy is this: does use of different ^qanalytic machineries^q entail a difference of ^qview^q? I put my q's in here to remind us that both ^awanalytic machineries^aw and ^awview^aw have to do with little-explored territories though they are surrounded by the most debatable land in ^qthe Western philosophic tradition^q. Current use of most of this machinery is erratic: at a popular level it cares little which of the above words are employed; more sophisticated use varies from one philosophical school to the next. (Richards 1953: 256)

An example of the use of this notation. If it actually be used it is advisable to make AutoKey "phrases" out of all the <sup>s.

Translation theory - over and above the aid it may afford the translator - has thus a peculiar duty toward man's self-completiono to use a concept which seeems to be suggestively common to the Chinese and the Western traditions. We are not weather vanes, I; we are not filing systems, II; we are not even agonies or delights only, III; we are not litmus paper, IV, or servo-mechanisms, V. We are guardians, VI, and subject therefore to the paradox of government: that we must derive our powers, in one way or another, from the very forces which we have to do our best to control. Translation theory has not only to work for better mutual comprehension between users of diverse tongues; more central still in its purposing is a more complete viewing of itself and of the Comprehending which it should serve. (Richards 1953: 261)

A noble sentiment.

See Cybernetics: Transactions of the Eighth Conference (Josiah Macy, Jr., Foundation, 1951); I. A. Richards, "Communication between Men," pp. 54-60; (Richards 1953: 262)

I wish I could. I can only hope that one of these days archive.org digitalizes it.