[Mova](https://md.mova.elamur.ai/free/en/) · [this page in HTML](https://mova.elamur.ai/free/en/docs/papers/postgate-1922-translation-translations-theory-practice/)

[home](https://md.mova.elamur.ai/free/en/) / [docs](https://md.mova.elamur.ai/free/en/docs/) / [papers](https://md.mova.elamur.ai/free/en/docs/papers/) / [Machine translation](https://md.mova.elamur.ai/free/en/docs/papers/mtrans/) / Translation and Translations: Theory and Practice

# Translation and Translations: Theory and Practice

**Authors:** J. P. Postgate · **Year:** 1922 · **Venue:** book (London: G. Bell and Sons) **Link:** [archive.org/details/translationtrans00postiala](https://archive.org/details/translationtrans00postiala) · **Licence:** [public domain](https://creativecommons.org/publicdomain/mark/1.0/) (the author died in 1926)

classic

- [Publisher's page](https://archive.org/details/translationtrans00postiala)
- [PDF](https://archive.org/download/translationtrans00postiala/translationtrans00postiala.pdf)

**Authors (1)**

1. J. P. Postgate

Citations and author metrics from OpenAlex (); authors with the same name are not always told apart.

## What it is about

A classical scholar's handbook of translation, in three chapters of principles, practice and verse, with his own renderings as specimens. Postgate makes fidelity the first merit of a translation proper. He splits translation by aim: retrospective (receptive) translation looks back to the author, and prospective translation looks forward to the reader. For the first, he says, the judge is the expert who knows both languages, never the general reader, for whom a translation is simply an original. Under "Degrees of correspondence" he borrows Schopenhauer's circles. Words of two languages may coincide (technical terms such as *triangle* and *Dreieck*), not correspond at all (*augur*, *premier*), overlap in part, or correspond only in appearance. The last are illusory correspondences such as *lurid* and *luridus*: same letters, same origin, almost no shared meaning. Later sections argue that translations age faster than originals, because translators import their own age. They say archaism is right only where the original was archaic or primitive for its own readers. Butcher and Lang's view that Biblical English suits epic Greek is cited with approval, but Demosthenes must be put into contemporary speech. A "principle of compensation" allows an effect that cannot be kept in one place to be rendered in another.

## Ideas

- I1. Illusory correspondences inside English. The interlingua lexicon anchors two Bibles on equal lemmas (the longest common subsequence in `meta.rs`). Equal lemmas across centuries are Postgate's illusory correspondences: KJV *let* (hinder), *prevent* (go before), *conversation* (conduct), *quick* (alive), *meat* (food), *suffer* (allow). Build a small test list of such pairs. Measure how often the reader anchors on them, and whether WEB's word at that place is a different concept. Anchors that fail it should be cut by context, not by lemma.
- I2. Score each concept merge by degree of correspondence. Two members coincide when their neighbour distributions in the gaps mostly agree, overlap in part when they share some contexts, and are illusory when they share only the form. Report the merge counts in the three classes with the reader and encoder bits, as a diagnostic for open problem 2.
- I3. Archaism by the age of the original. Postgate allows archaism only where the source was archaic for its own readers. In the translator this means a style delta's strength may depend on the source genre: Hebrew poetry and archaic songs (Genesis 49, Exodus 15, Judges 5) versus narrative prose. Measure whether the KJV delta's gain differs between these books.
- I4. Deltas are dated. KJV (1611/1769), ASV (1901), JPS (1917) and WEB (2000) lie on a time axis. Postgate's claim that translations age with their era predicts that the distance between deltas, measured as the number and the bits of overrides that differ, grows with the years between them. A cheap check on the deltas already learned.
- I5. Compensation as a verse-level measure. An emphasis or an antithesis may move within a verse. Next to the position-wise residual, count each class (negations, emphatic particles, repetitions) per verse in source and output, so a moved effect is not scored as one drop and one insert.

## For Mova

I1 and I2 give the lexicon in `meta.rs` its most concrete test set: false friends across centuries of English. I3 and I4 are reports over the deltas in `profile.rs`. I5 is a second view for `measure.rs`.

Topics: [Machine translation](https://md.mova.elamur.ai/free/en/docs/papers/topics/machine-translation/)

## Cite

The paper itself, not this card. BibTeX from the library's record of the work.

```
@book{postgate1922translation,
  title = {{Translation and Translations: Theory and Practice}},
  author = {Postgate, J. P.},
  year = {1922},
  url = {https://archive.org/details/translationtrans00postiala}
}
```

- [← previous A Call for Clarity in Reporting BLEU Scores](https://md.mova.elamur.ai/free/en/docs/papers/post-2018-call-clarity-reporting-bleu-scores/)
- [next → Aligning English Strings with Abstract Meaning Representation Graphs](https://md.mova.elamur.ai/free/en/docs/papers/pourdamghani-2014-aligning-english-strings-abstract-meaning/)
- [↑ back to Machine translation](https://md.mova.elamur.ai/free/en/docs/papers/mtrans/)