Bible translation is a complex task of studying the vocabulary, grammar and syntax of both the original language (Hebrew in the Old Testament, Greek in the New Testament, and Aramaic in a few chapters in the Old Testament) and the destination language (e.g., English). Part of the complexity is due to the fact that words in one language often do not have an exact equivalent in another language. Two major translation philosophies seek to overcome this difficulty.

Formal equivalency seeks a word-for-word literal equivalency as often as possible. However, this sometimes makes no sense in the destination language because of differences between languages in sentence structure and grammar. Dynamic equivalency seeks to communicate the exact meaning of the words. However, this sometimes results in an interpretive translation which is more subject to the interpreter's bias.

No translation completely conforms to either method. To calculate how closely various translations conform to either method, a deviation test has been developed. This test has shown the KJV, NKJV and NASB to be fairly similar literal translations. The NIV falls in the free translation category with the goal of translating ideas, and the Living Bible and Phillips Modern English Bible are paraphrases. The American Standard Version of 1901 is the most literal. Literal translations generally make more reliable study tools.

Attempting to capture the nuances of the source language sometimes results in awkward English. The ideal is to eliminate distortion and awkwardness without sacrificing faithfulness to the original text as much as possible.

The Deviation Test

The deviation test first assigns consecutive numbers to the original Hebrew or Greek words. Each word is translated into its nearest English equivalent, keeping the Greek or Hebrew word order intact. The English word order is then adjusted along with whatever other changes are necessary to achieve readable English. The original sequential number remains with the original words. This produces the closest accurate and meaningful English equivalent of the Hebrew or Greek text which becomes the norm for comparing other translations.

English translations are then compared with this translation by noting five kinds of deviations: changes in word order, omissions from the text, lexical alterations, syntactical alterations and additions to the text. Each deviation is assigned a numerical value depending on the kind and degree of difference between the closest equivalent and the version under investigation. The values are totaled to establish the deviation value for the section. This number is adjusted to reflect the deviation value per one hundred words. The process is repeated for other samplings, and the values are averaged to obtain a single deviation for the whole book. Then versions can be compared. (Thomas, Robert L., An Introductory Guide for Choosing English Bible Translations, 1988, p. 52)