Starting the translation straight away looks fast but usually slows things down: the questions that come up as the text progresses are asked one by one, and every question leads to waiting. Reading the text from the start allows these questions to be asked all together. In addition, the source text's own problems come to light at the earliest point.
The purpose of the read-through
The first reading is not for translating but for getting to know the text and seeing the problems. This reading lays the ground for all later decisions. Its duration is included in the plan.
Marking ambiguous wording
An ambiguity in the source text is not removed by translation; it is only carried into the target language. Ambiguous sentences are marked and asked about. They are not translated by guesswork.
Detecting contradictions
If the text contradicts itself, this is noticed early and reported. A contradiction noticed during translation stops the job. Early detection saves time.
Missing sections
Sections that are referred to but are not in the file are listed. The omission affects the scope. The client is told.
Term extraction
During the read-through, recurring and disputed terms are put on a list. This list is the input for term preparation. The work starts early.
Identifying abbreviations
The abbreviations in the text are collected and those whose expansion is not known are marked. When they are asked about together, the answer comes quickly. Asking one by one is slow.
Style and register
Whether the text is in official, technical or promotional language is determined. The style of the target text is built accordingly. The decision is made at the outset.
Estimating the target reader
Whom the text was written for supports the decision on style. An expert reader and an end user require different language. The information is verified with the client.
Repeated sections
Patterns that recur within the text are detected and resolved once. This provides both speed and consistency. The memory supports this work.
Numeric density
Sections dense with numbers and tables are marked and planned for additional review. These sections take as much time as text. The plan is made accordingly.
Image dependencies
The images the text refers to are identified and it is verified that they are accessible. A text without its images is understood incompletely. Anything missing is reported.
Spelling and typesetting errors
Obvious errors in the source text are noted and reported to the client. They are not silently corrected in the translation. The decision is the client's.
Scanning measures and units
Inconsistencies in the use of units are noticed early. The inconsistency is asked about before it is carried into the translation. Clarity is ensured.
Singling out binding wording
Sentences that create an obligation are marked and become an area of special attention. These sentences are not simplified. Marking them makes review easier.
Drawing up the list of questions
At the end of the read-through all the questions are gathered in a single list and put to the client together. Scattered questions delay the answer. The list method is efficient.
Prioritising the problems
Some questions must be answered before work can go on; others can be resolved later. The distinction is made. Waiting time is reduced.
Sharing the assessment note
The observations are passed on to the client in a short note. The note both provides transparency and manages expectations. A record is kept.
The next step
Once the text is known, the terminology infrastructure is set up. A job started without terms is corrected in bulk afterwards. The order is kept.