Simple explanation
What does editing for precision mean at C2 level? At C2 level, editing goes beyond correcting grammar. A sentence can be grammatically correct yet still be vague, unnecessarily long, logically uneven, or open to multiple interpretations. Precision editing aligns grammatical structure, vocabulary, information structure, and the strength of a claim with the exact meaning the writer intends. Precision, clarity, and concision are different goals Precision concerns exact meaning. Clarity concerns how easily that meaning can be processed. Concision concerns removing language that adds no useful information. A shorter sentence is not automatically better. If shortening removes a necessary limitation, contrast, condition, or degree of certainty, the result may be concise but less accurate. Recognising syntactic ambiguity Ambiguity often comes from the position of adverbs, prepositional phrases, relative clauses, and participial phrases. A modifier should normally appear close to the element it modifies. When it is positioned too far away, readers may attach it to a different part of the sentence and derive an unintended meaning. Making reference explicit Pronouns such as it, this, that, they, he, she, and which work efficiently only when their referents are immediately recoverable. If two nouns could plausibly be the antecedent, repeating the relevant noun or restructuring the sentence is usually preferable to preserving elegant but ambiguous reference. Reducing wordiness without losing meaning Long expressions can often be replaced with shorter equivalents. Due to the fact that can often become because, in order to can become to, and make a decision can become decide. However, editing should not remove words merely because they are optional grammatically. Qualifiers such as approximately, primarily, potentially, and only may carry essential meaning. Using nominalisation selectively Nominalisation is not inherently poor style. It is common and useful in academic and professional writing, especially when a process or concept itself is the topic. Problems arise when strings of abstract nouns hide agency or make a simple action unnecessarily heavy. Conduct an evaluation of can often become evaluate, but evaluation may be appropriate when the evaluation itself is being discussed. Using parallel structure to clarify relationships Items that perform the same grammatical and conceptual role should usually have matching forms. A sequence such as to reduce costs, improving reliability, and the expansion of access forces the reader to process three unequal structures. Reduce costs, improve reliability, and expand access makes the three goals immediately comparable. Controlling the strength of claims Words such as prove, demonstrate, and cause make strong claims. Suggest, indicate, be associated with, and contribute to are more cautious. These differences are not merely stylistic: they change the epistemic strength of the statement. Advanced editing requires matching linguistic certainty to what the evidence actually supports. Controlling negation and focus The placement of not, only, even, almost, just, and merely can substantially change meaning. We did not interview all participants does not necessarily mean that nobody was interviewed. If the intended meaning is that some participants were interviewed but others were not, it is better to state that explicitly. Editing questions for precision Questions can also be grammatically correct but too broad to answer precisely. Why is this study better? leaves the comparison criterion unspecified. A stronger version identifies the dimension and comparison point, such as Which aspects of the study design make its causal conclusions more reliable than those of the previous study?
Grammar formula
Reduce unnecessary nominalisation
make/conduct/carry out + nominalisation → precise verbThe committee conducted an evaluation of the proposal. → The committee evaluated the proposal.
Resolve ambiguous reference
ambiguous pronoun → explicit noun or restructured clauseWhen Maya spoke to Elena, she was frustrated. → Maya was frustrated when she spoke to Elena.
Position a modifier precisely
modifier + intended target in close proximityShe almost reviewed every file. → She reviewed almost every file.
Reduce a wordy expression
wordy phrase → concise equivalentdue to the fact that → because
Create parallel structure
X form + Y form + Z formThe role requires analysing data, interpreting results, and communicating findings.
Calibrate a claim
strong claim → evidence-appropriate claimThe results prove that X causes Y. → The results suggest that X may contribute to Y.
Clarify negative scope
implicit scope of not → explicit contrast or quantificationWe did not interview all participants. → We interviewed some, but not all, participants.
Correct an illogical comparison
property of A + comparison + corresponding property of BThe accuracy of Model A is higher than that of Model B.
Core rules
- 1
Place modifiers close to what they modify
Position adverbs, participial phrases, and prepositional phrases so that their relationship to the intended word or clause is obvious. Distance between a modifier and its target increases the risk of unintended attachment.
- 2
Replace ambiguous pronouns with explicit references
If a pronoun has more than one plausible antecedent, repeat the relevant noun or restructure the sentence. Controlled repetition is preferable to unclear reference.
- 3
Remove expressions that add little information
Phrases such as it is important to note that, the fact of the matter is that, and at this point in time can often be removed or shortened unless they serve a deliberate rhetorical function.
- 4
Prefer precise verbs when a heavy verb-noun phrase adds nothing
Expressions such as make a decision, conduct an analysis, and carry out an investigation can often become decide, analyse, and investigate. The simpler verb usually makes agency and action more visible.
- 5
Keep coordinated elements parallel
Items in lists, contrasts, and coordinated structures should normally have matching grammatical forms when they perform equivalent functions. Parallelism improves both readability and logical transparency.
- 6
Control the scope of only and not
Place only as close as possible to the element it restricts. Make the scope of not equally explicit. Small changes in position can change who, what, or which action is being limited or negated.
- 7
Match claim strength to evidence strength
Do not use prove, cause, or demonstrate automatically when the evidence only supports association, probability, or a tentative interpretation. Choose reporting verbs and modal expressions that reflect the actual level of certainty.
- 8
Do not remove every nominalisation
Keep nominalisations when the process, concept, or result itself is the topic. Revise them mainly when a direct verb expresses the same content more clearly and efficiently.
- 9
Make comparisons logically symmetrical
Compare like with like. The accuracy of Model A is higher than Model B incorrectly compares a property with an entity. The accuracy of Model A is higher than that of Model B creates a valid comparison.
- 10
Protect necessary qualifications during editing
Words such as approximately, partly, potentially, mainly, and under these conditions may be central to accuracy. Never delete a qualifier simply to make a sentence shorter.
Useful examples
We interviewed some, but not all, participants.
We interviewed several participants, but the entire group was not included.
The revised sentence removes the scope ambiguity that can occur in We did not interview all participants.
She reviewed almost every application before noon.
Before noon, she had checked nearly all of the applications.
Almost directly modifies every application, making the intended quantity clear.
Daniel was already concerned when he spoke to Marcus.
Daniel was worried before or during his conversation with Marcus.
Naming Daniel explicitly removes the ambiguity that would arise with he.
The revised procedure reduced processing time by approximately 18 percent.
The new version of the procedure cut processing time by about 18 percent.
Approximately is meaningful because it signals that the figure is not presented as exact.
The researchers evaluated the model under three conditions.
The researchers assessed the model in three different conditions.
Evaluated is more direct than conducted an evaluation of without changing the core meaning.
The policy aims to reduce costs, improve reliability, and expand access.
The policy has three goals: lower costs, better reliability, and broader access.
The three verbs create a clear parallel structure.
The data suggest an association between sleep duration and memory performance.
The evidence indicates that sleep duration and memory performance may be related.
Suggest an association avoids claiming a causal relationship that the evidence may not support.
Only senior editors can approve the final version.
Senior editors are the only people permitted to approve the final version.
Only modifies senior editors, so the restriction applies to who can give approval.
Senior editors can approve only the final version.
Senior editors are allowed to approve the final version, but not other versions.
Here only restricts what can be approved rather than who may approve it.
After reviewing the evidence, the panel revised its conclusion.
The panel examined the evidence and then changed its conclusion.
The implied subject of reviewing is correctly matched with the panel.
The device, which was tested independently, met all three safety requirements.
The device passed all three safety requirements, and it had also been independently tested.
The comma-delimited relative clause adds supplementary rather than identifying information.
The devices that passed the initial test were examined again.
Only the devices that succeeded in the first test were tested again.
The relative clause restricts which devices are being discussed.
Because the deadline had changed, the team revised the schedule.
The team changed the schedule because the deadline was different.
Because is more concise than due to the fact that while preserving the causal relationship.
The report distinguishes between temporary fluctuations and persistent changes.
The report clearly separates short-term variation from lasting change.
Distinguishes expresses a more precise relationship than a vague phrase such as talks about differences.
What evidence supports the claim that the intervention improved retention?
Which evidence shows that the intervention may have improved retention?
The question specifies exactly what kind of support the respondent should provide.
The change may improve efficiency, but its long-term effects remain uncertain.
The change could make the process more efficient, although its future effects are still unclear.
May and remain uncertain carefully control the level of certainty.
The accuracy of Model A is higher than that of Model B.
Model A is more accurate than Model B.
That of ensures that accuracy is compared with accuracy rather than with the model itself.
We currently lack sufficient evidence to draw a firm conclusion.
At present, there is not enough evidence for a definite conclusion.
The sentence is concise while preserving the important limitation on certainty.
Common mistakes
- ❌ After reviewing the report, several errors were discovered. ✅ After reviewing the report, the editor discovered several errors. The introductory phrase requires a logical subject that could actually review the report. Several errors cannot perform that action, so the original contains a dangling modifier.
- ❌ When Daniel spoke to Marcus, he seemed unusually nervous. ✅ Daniel seemed unusually nervous when he spoke to Marcus. He can grammatically refer to either Daniel or Marcus. Repeating Daniel makes the intended reference explicit.
- ❌ The study completely proves that social media causes anxiety. ✅ The study provides evidence that social media use may contribute to anxiety. Proves and causes express very strong certainty and causality. If the research design does not establish causation, the revised wording is more accurate.
- ❌ The company made a decision to discontinue the service. ✅ The company decided to discontinue the service. Made a decision to is an unnecessarily heavy nominalised construction when decided communicates the same action directly.
- ❌ The position requires analysing data, to prepare reports, and effective communication. ✅ The position requires analysing data, preparing reports, and communicating effectively. The original list mixes grammatical forms. The revision uses three parallel -ing structures.
- ❌ She only explained the second problem to the new students. ✅ She explained only the second problem to the new students. If the intended meaning is that no problem except the second one was explained, only should appear immediately before the second problem.
- ❌ At this point in time, we are currently unable to reach a final conclusion. ✅ We are currently unable to reach a final conclusion. At this point in time and currently communicate essentially the same temporal information, creating redundancy.
- ❌ The reason why the experiment failed was because the sensor malfunctioned. ✅ The experiment failed because the sensor malfunctioned. The reason why and because both signal causation. The direct version expresses the same relationship more efficiently.
- ❌ The accuracy of Model A is higher than Model B. ✅ The accuracy of Model A is higher than that of Model B. The original compares the accuracy of one model with another model rather than with that model's accuracy.
- ❌ The results show a possible potential improvement. ✅ The results show a possible improvement. Possible and potential perform essentially the same semantic function here, so using both is redundant.
Learning tips
- Ask what exact claim each sentence is making before editing its wording.
- Do not shorten mechanically; remove only language that adds no necessary meaning or structural function.
- If a pronoun could refer to two different nouns, repeat the important noun rather than forcing the reader to guess.
- Check only, merely, even, almost, just, and not carefully because their position can alter semantic scope.
- Use prove, cause, and demonstrate only when the evidence genuinely supports that level of certainty.
- Read difficult sentences outside their paragraph to test whether references and logical relations remain clear without extra context.
- Do not treat nominalisation as an error category. Revise it only when a verbal structure expresses the same idea more clearly.
- Avoid unnecessary synonym substitution in technical writing when repeating the same term preserves conceptual consistency.
- Check that parallel list items are conceptually equivalent as well as grammatically parallel.
- After shortening a sentence, recheck all qualifiers, numbers, modal verbs, exceptions, and conditions to ensure that precision has not been lost.
Practise this topic
Interactive exercises for this topic are coming soon.
PlannedRelated grammar
Related dictionary words
Frequently asked questions
Does professional editing always make an English sentence shorter?
No. The goal is not maximum brevity but maximum precision with minimal unnecessary language. A longer sentence may be better if the additional wording clarifies a condition, limitation, contrast, reference, or degree of certainty.
How can I tell whether a pronoun is ambiguous?
Check whether the pronoun could grammatically and logically refer to more than one earlier noun. If two plausible antecedents exist, repeat the intended noun or restructure the sentence.
Are nominalisations bad style at C2 level?
No. Nominalisations are a normal feature of academic and formal English. They become problematic when they unnecessarily hide the agent, create long chains of abstract nouns, or make a simple action harder to process.
Why does the position of only matter so much?
Only is a focus particle, so its position influences what is restricted. Only John called, John only called, and John called only Mary do not express the same meaning.
Is repeating the same noun poor advanced writing?
Not necessarily. Repetition can be preferable when replacing the noun with a pronoun or approximate synonym would introduce ambiguity or conceptual inconsistency. Precision is more important than artificial lexical variety.
What is the difference between precision and concision?
Precision means expressing exactly the intended meaning. Concision means expressing the necessary meaning without unnecessary language. A sentence can be concise but imprecise, or precise but more verbose than necessary.
When are verbs such as prove and cause too strong?
They are too strong when the research design or evidence does not establish certainty or causation. With observational, correlational, or preliminary evidence, suggest, indicate, may contribute to, and be associated with may be more accurate.
Should I remove words such as approximately and potentially when editing for concision?
Only if they are genuinely unnecessary. Such words often define the scope, uncertainty, or accuracy of a claim. Removing them can make a sentence shorter while making the content less accurate.
Can a grammatically correct sentence still need major editing?
Yes. Grammar alone does not guarantee clear reference, logical comparison, appropriate claim strength, efficient information structure, or precise modifier placement. These are central concerns in advanced editing.
How can I check whether a revised sentence is truly unambiguous?
Read it with as little surrounding context as possible and ask whether its references, comparisons, negation, logical relationships, and degree of certainty remain clear. If the intended interpretation still depends on guesswork, revise again.