How was the review done?
The reviewed text of a quite precise translation might look like this:
Original text: This is a short text.
Translation A: This was₁ a *new³₁ small¹ ·TEXT³₁.
The verb “was” receives 1 deviation point for form-faithfulness, because the tense is changed from present to past.
The added word “new” and the omitted word “text” receive 3 deviation points for literal accuracy, and 1 deviation point for
form-faithfulness. (Adding or omitting a word is always considered a formal deviation.) The word “small” receives 1 deviation point for
literal accuracy, because it differs a bit from the original meaning “short”.
When the translation is as accurate as this, it is quite easy to understand how the translation differs from the original text, and what is the reason
for each deviation score. Majority of Bible translations are at least as accurate as this, most of the time.
However, some translations are paraphrases, which can differ much more from the original text. Also a generally accurate Bible can translate some
expression idiomatically, so that the form and literal meaning of the expression changes a lot. In that case the deviation markings in the review
can look much more complex and difficult to understand:
Original text: This is a short text.
Translation B: Briefly passing as life itself.
This “translation” does not contain even one same word as the original text. How would we review this translation? The roughest method
would be to mark every word of the original text as an omitted word, and every word of the translation as an added word. Then the review would look like this:
Translation B: ·THIS¹₁ ·IS¹₁. ·A ·SHORT³₁ ·TEXT³₁. *Briefly³₁ *passing³₁ *as¹₁ *life³₁ *itself¹₁.
(I generally don’t give any deviation points for indefinite and definite articles, in any scenario, because their usage differs so much between
hebrew, greek and english languages.)
With this reviewing method, the total deviation score of this translation would be 19 points for literal accuracy, and 9 points for form-faithfulness.
This is more than twice the score that would be given for simply omitting the whole verse (literal accuracy 8, form-faithfulness 4), because each
mistake is punished twice: once for omitting something, and then also for adding something.
Even the most liberal paraphrases in Bible translations are usually based on (or somehow inspired by) the words that are found in the original text.
It would be difficult to keep the review scores fair and comparable with each other, if a translated word can be punished once with 3 points, if
the meaning is completely different, but twice with total 6 points, if the reviewer decides that there is no connection between the original word
and the translated word: one word is omitted, and a different word is added. In that case a subjective decision by the reviewer could double the
penalty that is given to the translation, as the reviewer makes this decision: “Is this word in the translation intended as a counterpart for
that other word in the source text, or is it not?”
To minimize the impact that such subjective decisions could have on the results of the review process, I rarely use a double penalty for omission
and addition. Such scenarios are rare, in which I would not find some word in the translation to act in the role of a “mistranslated”
counterpart for each word in the source text. If an obvious translation is not found for a word, and if some loose extra words remain without detected
counterparts, I can use very different words as “mistranslated” counterparts of each other. This policy results in quite lenient deviation
scores for paraphrases, as a double penalty of omission and addition is rarely enforced. But paraphrases fall far behind the precise translations in
the total scores in any case, and this policy is fairly the same for all translations. This policy minimizes the impact which subjective decisions might
otherwise have on the total deviation scores.
With my typical reviewing method of paraphrases, the review of this translation would look as is shown below. The total deviation score is 8 points for
literal accuracy, and 6 points for form-faithfulness = total 14 points, which is half less than the 19 + 9 = 28 points that we got afore, with the method
that punished omissions and additions separately.
Original text: This is a short text.
Translation B: ₁Briefly₁ passing³₁ *as¹₁ life³ ₁itself¹₁.
As we compare this paraphrase with the original text, we decide that “briefly” is intended as a translation for “short”.
The meaning “brief” vs. “short” is so close that we don’t give any deviation points for literal accuracy.
We give 1 deviation point for form-faithfulness, because the form has changed from “brief” into “briefly”. And 1 point
for structural deviation, because this word should refer to a noun, “text”, but there is no such noun anywhere around. (We will find
a counterpart for that noun a bit later.)
Both phrases contain one verb, so we decide that the verbs are counterparts of each other. The word “passing” receives 3 deviation
points for literal accuracy, because the meaning “pass” differs quite much from the original “be”. (In the first reviewing
method the omitted word ·IS¹₁ was considered a minor word, which is worth 1 point for literal accuracy. But the added word *passing³₁ was considered
a main word, worth 3 points. A similar change in deviation points is applied also in this latter reviewing method, for the same reason.) We give 1
deviation point for form-faithfulness, as the verb form has changed from “passes” into “passing”.
Both phrases contain one noun, so we decide that the nouns are counterparts of each other. The word “life” receives 3 deviation
points for literal accuracy, because this meaning is totally different than the original meaning “text”. We give no deviation
points for form-faithfulness, as both forms are in the same basic form (nominative) and in singular.
Both phrases contain one pronoun, so we decide that the pronouns are counterparts of each other. The word “itself” receives 1 deviation
point for literal accuracy, because the original meaning “this” has changed into “it”. (This minor word is not considered
worth more than 1 point.) We give 1 deviation point for form-faithfulness, as the basic form “it” has changed into
“itself”. And 1 point for structural deviation, because this word should be a stand-alone subject of the sentence, but instead it has
become a second violin for the new subject “life”. (We don’t punish the changed subject twice.)
Now as we have found a counterpart for every other word (except the indefinite article), we conclude that “as” is an added minor word,
which receives 1 deviation point for literal accuracy, and 1 point for form-faithfulness.
This is how the logical comparison of a paraphrase and the original text can look like, in the worst case. Some scenarios in the reviewed Bible
translations are yet much more complex than this example. In Other Laws 23:10 the translations EASYENB and EBDEAF (among a few others) contain
complex changes in the phrase structure. It can be difficult for the reader to get a complete understanding, what is the reason and meaning of
each deviation marking in those verses.
You should be able to decipher every detail in 5 minutes, if you deem it necessary, if you copy into a text processor the reviewed
text of a Bible translation, and the deviation scores of the whole verse text (which are given at the beginning of each verse on this page), in this way:
·IF¹₁ ·WILL₀ ·BE¹₁ ·AMONG₁ ·YOU³₁ ·A ·MAN³₁ ·WHO¹₁ ·WILL₀ ·NOT³₁ ·BE³₁
·CLEAN³₁, ·FROM₁ ·INCIDENT³₁
·OF₁ ·NIGHT₁³, ·THEN¹₁
·HE³₁ ·SHOULD⁰₀ ·DEPART³₁ ·TO₀ ·OUTSIDE¹₁ ·OF₀ ·THE ·CAMP³₁. ·HE³₁ ·SHOULD⁰₀
·NOT³₁ ·COME³₁ ·TO₁ ·MIDST¹₁ ·OF₀ ·THE ·CAMP³₁.
[EBDEAF] If ·WILL₀ ·BE¹₁ ·AMONG₁ ·YOU³₁ ·FROM₁ *something¹₁ happens₁
to₁ ₀a ₁man during₁ the night *that¹₁ ₁makes³₁ ₁him¹₁
₁impure₁, then⁰₀ he must⁰₀ go ·AWAY¹
·TO₀ out₁ of the camp. He may⁰₀ not stay³ ·TO₁ in ·MIDST⁰₁ ·OF₀ the camp.
If you follow the same routine as was used afore for “translation B”, and you mark with attention colour each pair of words, which
are apparent counterparts of each other, eventually you should find a pair and a logical explanation for every deviation marking
in the verse text. If you are unsure, what the reviewer has interpreted to be the counterpart for “during”,
all deviation markings are listed in the documentation of the verse, further below on this page. There you will find the
information that is shown below, which tells that the reviewer has interpreted “during” as a counterpart for “·OF₁”:
| ·OF₁ / during₁ / ₁during₁ / at₁ / by₁ / in₁ |
If the verse has a lot of text and many phrases, usually I search pairs for “orphan” words without an obvious counterpart from within
the same phrase only. I consider words in more remote locations only if they are “orphan” in their own context, and have a meaning that
obviously matches a meaning that is missing from another part of the verse. Translators of paraphrases sometimes do this that they move a word into
a different phrase and a different context. A connection between the original word and a translated word somewhere far away, which is detected by this
method, can often be what the translators actually intended, when they created their freely flowing paraphrase.
Many english-speakers can have difficulties in perceiving the minor details of form-faithfulness in participles, because
they are very ambiguous in english language. In the following three examples the word “running” is written and pronounced in the
same way, but its grammatical meaning is slightly different in each case: 1) He is running. 2) He is a running one. 3) Running in haste, he reached the bus.
In some languages (such as my native language, finnish) the difference between these is obvious, because the form is different in each case. If someone
uses the wrong form, everyone will understand that something is not correct in the sentence. But english-speakers can have difficulties in perceiving
what the difference is between these grammatical scenarios, because the form “running” is casually the same in each case.
The expression “voice of a shouting one” in Loukas 3:4 is a scenario where I have given a deviation point for form-faithfulness (and another
point for a structural deviation) to many translations, which use the present continuous tense “is shouting”, or more simply “shouting”,
while I accept only the participle meanings “shouter”, “shouting one” or “one shouting” as formally identical with the greek
text.
Such translations as “a voice is shouting” or “a shouting voice” have a structural deviation in the sentence, because in greek text the
voice is not shouting. We know this from two details: The genitive “of the shouting one” shows that the voice and the shouter are different things,
so that the voice belongs to the shouter. Also the gender of the word “shouting” is masculine in greek, while “voice” is feminine.
Therefore the masculine “shouting one” must refer to something else than the feminine “voice”. These two details make it
grammatically obvious that the one who is shouting is someone else than the voice itself.
Each word in a sentence has a specific grammatical role. I have not required in this review that the sentence must remain exactly identical with the original. I pay
systematically attention only to the most obvious relations between word pairs, for example if a different word in the sentence becomes the subject of a verb
(not merely the same but mistranslated word), or if an adjective refers to a different main word than in the original text. These are significant changes
in the general thought, if an action is ascribed to a different doer (who did not do it), or if an adjective describes the wrong target.
Structural deviations usually receive 1 penalty point only, because the impact on the general thought is rarely remarkable. If a voice is shouting, or if the voice
of someone is shouting, the difference in thought is small. It is the same person anyway. In rare cases I can give 3 points for a structural deviation, if the change
in meaning is significant, for example “A says to B” is changed into “B says to A”. I give 3 deviation points for changing the location of the
word “not” in Romans 8:12, so that the wrong word gets negated.
English verbs and other expressions that consist of two or more words are not self-evident cases, how to treat them in the review of literal accuracy or
form-faithfulness. Should the unit of several words be treated as if it were one word, or should the words be handled separately, as if they had no connection
with each other? The answer to this question is not always obvious.
In Loukas 3:4 I have accepted without deviation points the two-word expression “one₀ calling” as an equivalent for one greek word, which
could be translated with one word as “caller”, but whose meaning is more precisely “a calling one₀”. In the same verse I have accepted
such two-word expressions as “calling out₀” or “crying out₀” as equivalents for the verb “shouting”. When a greek
word is translated with more than one english word, I usually mark the additional words with 0 deviation points for form-faithfulness (if I accept the more
complex expression as necessary and caused by the english grammar).
However, I have not accepted such expressions as “*a *man²₁ who₁ is₀ shouting₁” as literally and formally precise equivalents for the concept
“shouter” or “a shouting one₀”. I have treated “*a *man²₁” (and some other similar expressions) as added words,
which are marked with asterisks. I have accepted “who₁ is₀ shouting₁” as a logical equivalent for the concept “a shouting one₀”,
so it does not get any deviation points for literal accuracy. However, I give two deviation points for form-faithfulness (treating this in the formal review
as a sentence of three words, not as a single concept), because what could be said with much fewer words (or even with one word) has been converted
into an unnecessarily complex statement. Unnecessary complexity is typical for paraphrases, not for translations that are regarded as “precise”.
I accept significantly increased complexity without deviation points only when it seems that a complex expression has been absolutely necessary in the
target language. Psalms 41:9 provides such a scenario. The hebrew idiom “a man of my peace” is quite meaningless in english as a literal translation.
Therefore I have made an exception to the general policy in this review, and I have accepted it without deviation points, if it has been translated idiomatically
as something like “a man in₀ peaceful₀ terms₀ with₀ me₀”. (This meaning is not a speculative blind guess, because the Tanakh contains five verses
with a similar idiom, and this meaning would conveniently fit into all of them.) How idioms are handled in the review is an aspect where some subjectivity cannot
be avoided — unless we require a literal translation in idioms strictly in all cases, also when the result would not be a very reasonable thought in english.
These principles of handling paraphrases in the review can seem complex, and also subjective to some extent. Not following these principles might make the review
even more subjective, or less meaningful. Paraphrases will get a score far below the most accurate translations in any case, so it is perhaps not
crucially important, if their deviation points are 800 % or 900 % higher than the points of very accurate Bibles. The most important thing is that paraphrases are
systematically treated in a similar way, so that their deviation scores are fairly comparable with each other.
Other aspects of the review process are explained on the About page.