Who speaks in a translation: author, tradition, and translator

Epigraph

I have therefore cast it into our flexible blank verse

Edwin Arnold, preface to The Song Celestial, phrase explaining his choice of metre. This is the translator's statement, not a line from the Bhagavad-Gita; the quoted phrase ends before the sentence's continuation.

What this article is about

When we read a translation, we encounter an author through another person's decisions. Someone chooses words, rearranges sentences, identifies speakers, explains an unfamiliar custom, and decides where difficulty should remain. An editor then adds a title, notes, and information about the edition. These contributions help us meet a work, but their voices can easily merge. The mistake becomes conspicuous when a confident translator's preface is quoted as an ancient text's own teaching. Here translation is considered as responsible mediation: the original statement belongs to its author, its accessible form involves the translator, and conclusions drawn from it belong to the reader. Selected arguments by Jerome and Schleiermacher, one preface by Edwin Arnold, and Locke's discussion of complex concepts help distinguish these responsibilities. They do not verify every translation against its original. The practical question is how to use a translated work confidently while keeping visible who is responsible for each claim.

I. A page's first person does not always belong to its author

Imagine an edition of an old dialogue. The cover bears a philosopher's name, a short preface precedes the text, and explanations appear in brackets. A reader finds “I preferred to preserve the familiar order” and records it as the author's admission. Yet it concerns the translator's work. The error arises before any argument about accuracy: the speaker has not been identified. A cover's name identifies the work without making every added line authorial. Even a carefully prepared edition requires us to distinguish its main text, introduction, notes, and commentary. This is ordinary accuracy of attribution, through which a quotation reaches its actual source and retains a role that another reader can inspect.

Start by locating the statement. A work's narrator, its historical author, and the writer of its introduction can each use the first person without becoming one person. A dialogue adds its characters. One participant's reply does not necessarily express the work's final position. Translation makes their speech accessible without removing this distribution of roles. To investigate a particular claim, note where it appears and what follows. A selected sentence might be a question, an objection, or a provisional concession. Its relationship to the subsequent argument may matter more than the solemn tone of an isolated line, whether that tone was retained or conveyed through the translator's own language.

This distinction gives readers room for selective agreement. We can value a translation while questioning its preface, accept a historical observation while disputing an editor's conclusion. We cannot first merge these positions and then declare an author inconsistent because other people's words collide. No suggestion is made that every edition misleads. The imagined example identifies a basic check: before debating content, recover the sentence's location and role. If the boundary is unclear, describe it as unresolved and consult another edition or information about its contributors. Acknowledged missing evidence is more accurate than confident attribution based on style, a book's age, or our impression of its profundity.

II. A mediator chooses form without acquiring every possible thought

Translation requires decisions. Even related languages arrange words and connect clauses differently. A translator may reorder a sentence to preserve a question, add an explanation to retain a distinction, or leave several possibilities open rather than present a guess as the text. This does not make replacing the original statement with a convenient thought acceptable. A formal choice must remain connected to what is being translated. Distinguish a change of means from a change of assertion: a rebuilt sentence can express the same relationship, while a familiar word can quietly introduce a different one. Neither literal proximity nor fluency is a sufficient check. The question concerns the relationship of meaning that survives.

In Letter57, Jerome defends his translation in a particular dispute. The selected sections describe its circumstances and his preference for conveying sense rather than reproducing words sequentially. His position also contains a qualification concerning Scripture and word order. This is a participant's polemical argument, not an impartial report establishing the disputed translation's quality. It can frame a question about fidelity to meaning; it cannot authorize every free rewriting through an ancient name. The sections were read in a historical English translation. The Greek and Latin texts involved in Jerome's controversy were not independently collated. That limit belongs to an honest comparison rather than being an incidental detail that can safely disappear.

Ordinary reading encounters a similar boundary between translation and adaptation. A difficult treatise recast as short contemporary formulas may be useful without reproducing every authorial step precisely. Readers benefit from knowing what has been shortened, joined, or explained. They can then choose an appropriate entrance and recognize when they need a fuller text. Naming the kind of work has a practical function: an adaptation can orient us, whereas a dispute about a particular formulation needs a different foundation. Trouble arises when a convenient entrance is presented as the entire work, leaving readers no route for distinguishing an author's assertion from what a mediator contributed to make it accessible.

III. Two directions of translation reveal different demands

In the selected passage of On the Different Methods of Translating, Schleiermacher distinguishes moving the reader toward the author from moving the author toward the reader. This is a historical normative account, not evidence that every contemporary translation falls neatly into two categories. Its value here concerns the direction of effort. Readers can be introduced to an unfamiliar way of speaking, or a foreign work can be presented as though it originally arose within a familiar linguistic setting. Each approach makes decisions about distance. Those decisions concern more than rare words: they shape which relationships seem ordinary, which require attention, and where readers notice another habit of expression.

Retained difficulty can protect a distinction, but difficulty is not always helpful. An obscure sentence may reflect the original, an unsuccessful translation, or inadequate explanation. Without examining those circumstances, discomfort cannot be treated as proof of fidelity. The reverse mistake is equally easy: suspecting a good translation simply because it reads smoothly. Schleiermacher's account helps discuss an intention; evaluating its realization still requires particular examples. When a translator identifies the foreign feature preserved, readers have something to examine. Merely saying that a text ought to be difficult provides a slogan without a way to assess any individual decision in its context or to tell useful distance from avoidable confusion.

Readers can vary their approach with their purpose. An introduction calls for coherence, investigating a term for consistent distinctions, performing a poem for sound, and a historical dispute for exact locations. No purpose automatically cancels the others, but they should not quietly replace one another. A translation made for reading aloud need not answer every question of scholarly collation. Extensive notes may help investigate a concept while requiring more time. Instead of ranking all translations at once, state the present task and ask which decisions serve it. Comparative judgment remains possible, while a reader's particular convenience is kept from becoming a universal test of fidelity to an author and every possible use of the work.

IV. Arnold's preface openly identifies a translator's choice

In the preface to The Song Celestial, his English version of the Bhagavad-Gita, Edwin Arnold explains choosing blank verse and, in certain passages, lyrical measures. He relates that decision to his view of Western listeners' reception of the original metre. What matters here is the explicit description of a change in form. His judgment about hearing is not treated as a measured finding, and his translation was not checked against Sanskrit. The preface shows how a translator presents his task and its result. It does not prove his chosen form is uniquely possible. Readers can consider that statement when reading the English rhythm while distinguishing the work's original verse form from this means of conveying it.

Arnold also evaluates earlier translations and offers historical suppositions about the work. Those judgments belong to his preface. In particular, its speculation about dating and possible connections with Christian texts is not accepted here as established contemporary history of the Bhagavad-Gita. It is mentioned solely to distinguish voices. One mediator can translate, compare, and interpret history, but success in one activity does not establish correctness in the others. The introduction cannot all become Krishna's speech; nor can disagreement with one historical hypothesis refute every line of the translation. Each conclusion needs its own object of examination and evidence appropriate to that object, with the status of the translator's additional claims preserved.

A quotation's independent circulation makes the distinction especially useful. A short preface phrase can be placed beneath an ancient work's title while losing its mediator's name. Readers then encounter a new claim of authorship. A minimal address prevents this: whose preface, which edition, which passage. This article's epigraph therefore names Arnold and identifies his explanation of metre. The phrase is not attributed to the main dialogue. Precision need not reduce a quotation's expressive force. It lets the words do their actual work: testify to a translator's choice rather than imply that the original author approved an English form through which the work is only now receiving another voice.

V. A tradition does not speak through every edition with one voice

“Tradition” can refer to a continuing reading practice, a history of commentary, a community, or a work's long transmission. Distinguishing these uses helps. One edition can draw on a tradition while selecting a particular direction within it. A familiar title does not guarantee access to all its positions. Readers need not study an entire history before opening a book, but should bound their conclusions. “This translation proposes” is more accurate than “the whole tradition teaches” when one translator provides the evidence. Likewise, one commentary does not become a vote across centuries. An identified interpretive lineage aids understanding; when it is unidentified, an assured presentation alone cannot establish a universal claim.

Tradition's mediation extends beyond specialized commentary. A familiar rendering can become so established that another seems distorted before its context is read. Familiarity and accuracy nevertheless provide different reasons. A new option should be examined for the distinction it retains and the understanding it enables; an older one deserves the same scrutiny. Neither age nor novelty settles the issue. When translators follow established terminology, ask whether it reveals the relevant thought or now needs explanation. When they depart, ask why. Discussion becomes concrete through comparisons of a particular passage's renderings, rather than turning one unfamiliar word into a symbolic contest over loyalty to an entire cultural inheritance and the worth of its readers.

Respect for tradition is compatible with distinguishing its internal voices. People can retain religious or philosophical commitments while recognizing several ways of reading. Comparison requires neither declaring these identical nor ranking people's dignity by their translation preferences. The subject remains a text and an argument. Questions about a community's internal norms require that community's own explanation rather than an outsider's assumption; historical questions require historical investigation. Those questions are not settled here. The selected evidence concerns mediators' work, not every line of Christian, Indian, or European thought. A limited map supports understanding when its boundary is visible and convenient answers are not drawn into territory for which evidence has not been gathered.

VI. Translation can join concepts that an author distinguished

In the selected passages of the Essay's third book, Locke discusses complex concepts whose combinations of features may lack matching counterparts across languages. His philosophical account is used here, not an experimental survey of every language. It suggests a practical caution: a familiar translated word does not guarantee familiar content. Translators choose an available sign; an authorial distinction has to be recovered from that sign's use in the argument. Different Russian words for one English term need not indicate disagreement about a doctrine. Conversely, one word translating two original terms may conceal a distinction that readers discover only through a note or consistent comparison of the terms' separate contexts and consequences.

Imagine a treatise distinguishing feeling, evaluation, and rule. A short paraphrase calls all three an “attitude.” Readers receive an intelligible general theme but lose the relationships used to separate cases. They might later accuse the author of confusion that actually arose during transmission. This is an invented example, not a judgment on the historical translations discussed here. It illustrates why a research question needs sufficiently detailed evidence. Combining categories can assist general orientation, but a conclusion about conceptual structure brings another obligation: identify what was combined. Fairness to a mediator consists in assessing the work against its stated task, distinguishing useful compression from unnoticed replacement, rather than assuming either inevitable error or perfect accuracy.

Small local comparisons help. Choose one passage, record its key word, and examine its relationship to nearby claims in two translations. There is no immediate need to declare the best version of the entire book. Identify what each choice makes visible or obscures: does a question remain a question, does negation survive, are an acting person and an evaluated action distinguished? Without access to the original, the conclusion stays limited: the translations present this passage differently, and final linguistic collation remains open. That is a substantive result. It neither hides uncertainty nor reduces comparison to taste, because it concerns an identifiable consequence of a verbal choice that another reader can discuss and inspect.

VII. A note should remain a note when copied

Editorial explanation can illuminate a difficult work if its attachment is visible. Brackets, footnotes, different type, or an initial statement establish a boundary. Copying often removes formatting, allowing added words to enter a quotation unnoticed. Recipients then cannot identify what occurred in the main text. Accurate transmission therefore preserves roles as well as letters. An explanatory addition should be marked as our own or the editor's; shortening should not create the impression of a continuous authorial sentence. No rules for every publishing format are laid down here. The concern is a condition through which the next reader can identify whose claim supports the discussion and which links were added along the way.

A note sometimes supplies context and sometimes interprets it. Identifying a dialogue's speaker can locate a voice; explaining why that voice expresses the work's central teaching requires an argument. Both activities are legitimate when their status is visible. “That is” deserves attention when followed by a new thought. Readers may hear it as a simple expansion even though it excludes several possible readings. Ask whether the passage supports that narrowing. If support is insufficient, retain an interpretation as a hypothesis rather than presenting it as the original's settled content. Marking roles does not prove an explanation. It lets readers see where evidence is needed and who has proposed a claim that needs it.

Digital circulation gives this discipline practical importance. A paragraph can pass through several retellings, keeping a work's title while losing its translator. Rapid transmission does not diminish the relevance of authorship. When a formulation carries an argument, keeping its specific edition and location helps. This is no demand to burden every ordinary message with a bibliography. A proportionate address is enough wherever attribution can change meaning. General impressions can be stated in our own voice; precise quotations require knowing what is quoted. This distinction also supports reader autonomy. Recipients can return to the same passage, compare alternatives, and judge an interpretation without having to rely solely on the authority of the name attached to it.

VIII. An inaccessible original changes a conclusion's strength

Many people read much of world literature through translation. Limited original-language access does not make that reading meaningless. It leaves some questions accessible while others need help. Readers can discuss an argument in their chosen edition, compare replies, and consider a proposed thought's consequences. Independently establishing that an original word had exactly one possible sense is harder. When these tasks remain distinct, translation supports substantive work without pretending to linguistic competence we do not possess. Saying “this translation expresses the distinction in this way” identifies the conclusion's real foundation. If a later question depends on the original form, the reader can specify what additional access is needed rather than vaguely dismissing all previous reading.

A second translation expands observation, but the number of versions does not establish truth by voting. Several editions may follow one earlier translation, differ only in editorial style, or represent different underlying texts. Treating them as independent witnesses requires information about their origins. Without it, comparison is still useful, but its conclusion changes: several presentations are available, while their independence is unestablished. This is more accurate than treating majority agreement as proof of authorial meaning. Consulting an expert likewise leaves a question about the reasons offered. A useful explanation reveals a decision and its limit; it does not require readers permanently to replace their own inquiry with obedience to the most confident interpretation.

Uncertainty matters especially before a strong claim about a person or community. If a translated phrase is used to attribute cruelty, superiority, or universal wisdom to a tradition, a lost qualification can change the entire argument. Extending the context or postponing generalization then makes sense. A less consequential conversation may require only marking a reading as provisional. Proportion means matching the depth of checking to what we intend to do with the text. It does not make unknown evidence known when a conclusion is convenient. Where an edition supplies only part of the foundation, a narrower conclusion can preserve the rest as open, avoiding both total distrust of translations and uncritical acceptance of every fluent formulation.

IX. Connect disagreement to a decision before inferring motives

Readers who find a translation unsuitable can first describe an observation. One version resolves an ambiguity that another retains; one repeats a term that another varies. These features can be discussed. A translator's hidden intention is a further claim. Without evidence, every difference should not become deception, censorship, or betrayal. Such an accusation often closes the very question worth investigating: what meaning changes because of the decision? A choice can be criticized forcefully and specifically without assigning a motive. Where editorial intervention is openly described, consider that account and examine whether the result serves its declared purpose, rather than building the discussion around a secret for which no evidence has been found.

Respect for labour does not require abandoning substantive criticism either. A translator can choose poorly, an editor can lose a boundary, and a reader can suggest a clearer explanation. Show the passage and its consequence. If negation disappears, identify the resulting assertion; if voices merge, say which; if a comment becomes a quotation, recover its status. Correction can remain local without reopening the entire book unnecessarily. This permits cooperation through substantial disagreement. A concrete observation allows a mediator to explain a reason, correct an error, or defend an alternative. Readers can assess the reply without turning the discussion into a competition over the participants' general cultural worth or their loyalty to a tradition.

Sometimes discussion leaves two defensible options serving different purposes. Preserve that difference and state what each can do. Not every disagreement ends with one version's removal. Yet acknowledging alternatives does not make every version equally supported. Some decisions can be justified, some refuted, and some left unresolved. Distinguishing them requires reading the particular text. None of the selected sources supplies a universal judge for all cases. They offer questions about meaning, direction, form, and complex concepts. A new passage still requires an answer built around its own task. This limitation preserves historical comparison's usefulness: older arguments guide attention without replacing present examination through an automatic appeal to a revered name.

X. Readers are responsible for moving from text to conclusion

After all the mediators comes another voice: ours. We relate a passage to experience, choose an example, generalize, or propose an action. The step may be justified without becoming authorial merely because it begins with a quotation. Distinguish “the text says,” “I understand this as,” and “I therefore propose.” Another person can then agree with a source while questioning its application. That possibility matters in conversation: disagreement need not become rejection of an entire work. Keeping the transition visible makes our position more examinable, because others can identify which parts depend on the text and where an additional judgment of our own enters the argument and needs its own reasons.

Reading this way is not an endless search for mistakes. We can trust conscientious mediators and use their work while retaining its foundations. Naming a translator, marking a preface, locating a passage, and recording a relevant qualification help us encounter a work. They cannot guarantee every decision's final accuracy, but let us revisit disputes. The same boundary is maintained here: Jerome, Schleiermacher, Arnold, and Locke were read in specified passages, while practical examples and conclusions are United Wisdom's editorial proposals. They are neither attributed to an entire tradition nor presented as a proven method for improving understanding. Readers can use the distinction of voices where helpful and assess its value against their own question.

A translated encounter remains an encounter when the mediator stays visible. We can learn about an author while understanding the route of access. Where that route includes compression, commentary, or a different verse form, its decisions can be named and discussed. Where original-language evidence is unavailable, the limit can remain without an invented final answer. The central result is a habit of placing a statement beside its actual speaker. Authors are then not assigned translators' choices, translators do not answer for every reader's inference, and our interpretations remain our own. Discussion becomes more precise, giving each participant a way to explain their contribution without disappearing into one undifferentiated voice of authority.

What this means in practice

Choose a short translated passage you already wanted to discuss. Locate its edition, translator, work, and passage reference. Identify where the main text ends and a preface, note, or editorial explanation begins. Leave unclear boundaries as questions. This exercise does not require reading a new treatise in full or proving an original correct: its task is to recover the voices on one selected page and identify what you actually have access to.

Record one assertion and label its role: a character's reply, an authorial argument, a translator's explanation, or your inference. If useful, compare the same passage in another edition. Identify one concrete difference and a possible consequence for understanding. Avoid inferring a mediator's intentions from one word. Without original-language access, describe your result as a comparison of translated versions and leave the source formulation's final assessment open for further evidence.

Tell another reader what you wish to discuss: the phrase's content, its rendering, or your application. Consider whether separating roles helps answer that question. More information may be necessary; the available context may suffice. This is an editorial practice proposal, not a method with established efficacy. Stop when the task is answered. Ordinary reading need not become a compulsory investigation into the provenance of every part of a book.

Key quotation

I render sense for sense and not word for word.

Jerome, Letter57, §5; historical English translation W.H. Fremantle, G. Lewis, W.G. Martley. The same section separately qualifies this position for Scripture and word order.

Connections between traditions

Comparison	Shared question	Significant difference
Jerome, Letter57 §5; Schleiermacher, two directions	How does mediation retain its connection with the original statement?	A defence of meaning in a specific controversy is not a general theory of moving reader and author.
Jerome §§2,5–6; Arnold's preface	How do translators explain decisions?	Polemical defence of a disputed rendering differs from openly describing a chosen verse form.
Jerome §5; Locke III.V.7–10	Why are matching words insufficient?	A norm of conveying meaning and a philosophical account of complex concepts answer different questions; neither proves the other.
Schleiermacher, two directions; Arnold's preface	How is distance between a work and readers considered?	A theoretical distinction does not establish Arnold's particular judgment about metre or replace linguistic collation.
Schleiermacher, selected passage; Locke III.V.7–10	What does crossing linguistic habits require?	The direction of a translator's effort differs from the question of concepts' unmatched combinations of features.
Arnold's preface; Locke III.V.7–10	Which decisions does another linguistic form reveal?	Announced changes of metre concern form; unmatched complex concepts concern a sign's content.

Primary sources

Jerome. Letter57 to Pammachius, §§2,5–6. W.H. Fremantle, G. Lewis, W.G. Martley, NPNF, Second Series, vol.6,1893; New Advent revision Kevin Knight. Translation circumstances and selected defence were actually read. Neither Latin nor the disputed rendering was independently collated.
Friedrich Schleiermacher. On the Different Methods of Translating, English translation Waltraud Bartscht. Selected passage from “Paraphrase seeks” through the two directions to “transforms him into one of them,” zero-index PDF pages3–5, actually read. No complete-work reading, German collation, or universally exhaustive classification claimed.
Edwin Arnold. The Song Celestial, preface; New York, Truslove, Hanson & Comba,1900, Gutenberg2388. Complete preface and edition credits actually read. Translator's statements about form are used; historical hypotheses are not adopted as established contemporary history. The main translation was not checked against Sanskrit.
John Locke. An Essay Concerning Humane Understanding, vol.2, III.V.7–10, Gutenberg10616, historical English edition. Selected complex-concept and cross-language context actually read. A philosophical question, not an empirical language survey or complete-book reading.

About the authors

Andrey Zarinsh is the author and founder of United Wisdom. He explores inner transformation, the unity of traditions, intercultural dialogue, and contemporary spiritual discipline. Instagram: @andrey.zarinsh.

Maria Evilkut is a co-author of United Wisdom. Alongside her research, she seeks to test these ideas in practice and observe how principles of unity and mutual understanding work in real life between people and cultures. Instagram: @ilkutki.