Translation by formal equivalence is predominantly oriented to preserving the linguistic form that the original has in the source language, endeavoring to imitate it in the succession of words, in the syntax and, as far as possible, in the sonority and phonology of the target language. Translation by dynamic equivalence, on the other hand, is focused on the pragmatic effect that a linguistic message in the base language tries to convey to the hearers-readers-recipients, and makes every effort to reproduce the same effect by means of recourse to the target language. These two manners of translating spring from different postures of the translator that convey, or try to convey, the original to the contemporary world, taking account of the literary tastes of the reader (dynamic equivalence), or, on the contrary, seek to reproduce as well as possible the tenor of the original in such a way that the reader feels transported to it and its period (formal equivalence). No doubt preference must be given to translation by dynamic equivalence, but avoiding the danger that transferring the original to the contemporary world might degenerate into an unfaithfulness to the text in order to accommodate the tastes of the reader.
On the other hand, it is not acceptable by any standard for us to understand translation as transferring the vocabulary of the source language to the target language, a defect which can be observed, for example, in practically the all English translations of the intertestamental literature in the work of R. H. Charles (The Apocrypha and Pseudepigrapha of the Old Testament in English, 2 vols. [Oxford, 1913]). Besides the avoidance of effort, in this tendency the reverential attitude to the ancient texts, considered sacred, has dominated, and the necessary formal freedom to communicate the meaning faithfully is lacking.
Though with some limitations, the translation of meaning from an ancient culture to another modern one is indeed possible, and consequently, the "translation" of the riches of an ancient text into a modern language, without there being too much lost in the process. A literary translation is indispensable if the text is rich in literary characteristics. The stylistic and literary study of the originals is necessary, and it is important that the translator, besides having a good command of his own language, does not only operate with the mentality of grammar and vocabulary. A translation based on dynamic equivalence is more faithful, i.e., in the precise passage of the concept of one language to another, though the literal meaning of the terms may vary. For instance, in Matt. 8:4 we find the prepositional phrase εἰς μαρτύριον αὐτοῖς, literally "for a testimony to them." In the translation God's Word to the Nations it is rendered "to show them that you are well." At Rom. 12:20 we find τοῦτο γὰρ ποιῶν ἄνθρακας πυρὸς σωρεύσεις ἐπὶ τὴν κεφαλὴν αὐτοῦ, literally "for by doing this you will heap coals of fire on their heads." But it could be rendered "you will make them go red with shame" as J. Mateos and L. Schökel have translated it. Similarly, a non-literal translation is appropriate for such words as σάρξ, σῶμα, ψυχή, πνεῦμα, etc., the purely mechanical translation of which would amount to a veritable betrayal of the original meaning.
Finally, it is essential when translating to study the deep structure of the texts, revealing the articulation of phrases and the logical framework, ideally presupposing a previous exegesis.
This is not to deny absolutely the possible didactic advantages of an interlinear translation in service (exclusively) of learning a foreign language and to help the beginner to grasp, analytically, the structure of the foreign language. But, once the first stage has been passed, this means of translation, which sacrifices the target language and in reality betrays the meaning of the language of origin, should be rejected.
*The above is taken, with only slight modification, from the first appendix in Antonio Piñero and Jesús Peláez, The Study of the New Testament: A Comprehensive Introduction, trans. David E. Orton and Paul Ellingworth, Tools for Biblical Studies 3 (Leiden: Deo, 2003), 515-517.