006-ADR: Language for Technical Documentation¶
Architecture Decision Record (ADR) describing the choice of language strategy for technical documentation.
1. State¶
- Author(s): Sentenz
- Date: 2025-10-14
- Status: Proposed
2. Context¶
Engineering, product, and compliance artifacts are produced by distributed teams and consumed by stakeholders across technical, legal, and operational domains. The choice of language strategy for authoring and publishing technical documentation is critical to ensuring clarity, maintainability, and accessibility.
-
Decision Drivers
-
Accuracy and Clarity
Technical documentation must be precise and unambiguous to avoid misinterpretation.
-
Maintainability
A single source of truth simplifies updates, version control, and reduces the risk of inconsistencies.
-
Accessibility
Documentation should be accessible to a global audience, balancing the need for localization with operational efficiency.
-
Speed of Iteration
Faster review and publication cycles are essential to keep documentation aligned with rapid software releases.
-
Cost Efficiency
Translation and localization efforts should be cost-effective, minimizing ongoing maintenance overhead.
-
AI Integration
Future-proofing documentation processes with AI tools for translation and quality assurance.
-
3. Decision¶
Adopt a Single Language (English) strategy for authoring and publishing technical documentation and requirements. The approach focuses on accuracy, maintainability, simpler CI automation, and speed of iteration while acknowledging the trade-offs in accessibility.
Where non-English access is required, translations will be provided selectively via controlled processes (human or AI-assisted with human review) prioritized by impact and compliance needs.
Note
The decision will be revisited as AI-assisted translation technologies mature and become more reliable without compromising the integrity of the documentation. Enterprise-grade solutions and governance frameworks that support production use will be evaluated for accuracy, security, and operational fit before adoption.
3.1. Single Language (English)¶
-
Rationale
-
Accuracy and Clarity
English is the most widely used language in technical fields, reducing the risk of misinterpretation.
-
Maintainability
A single language source simplifies updates and reduces the risk of inconsistencies.
-
Accessibility
A single-language strategy limits access for non-English speaking audiences.
-
Speed of Iteration
Faster review and publication cycles are achievable without the overhead of translation.
-
Cost Efficiency
Eliminates ongoing translation costs and reduces maintenance overhead.
-
AI Integration
No approved, enterprise-grade AI assistant for translation, terminology control, and QA.
-
4. Considered¶
4.1. Single Language (English)¶
Author and publish exclusively in English.
-
Pros
-
Simplified Maintenance
Simplifies maintenance and reduces risk of contradictory guidance.
-
Faster Review Cycles
Faster review cycles and lower cognitive load for maintainers.
-
Simplified CI Pipelines
Simplified documentation CI pipelines and automation.
-
Cost Elimination
Eliminates translation costs and delays.
-
Standards Alignment
Aligns with the majority of technical documentation standards.
-
-
Cons
-
Limited Accessibility
Limits accessibility for non-English speaking audiences.
-
Market Perception
Potential perception of reduced local market investment.
-
Additional Support Channels
May require additional support channels for non-English speakers.
-
4.2. Multi Language¶
Author in English and translate to multiple languages.
-
Pros
-
Global Accessibility
Increases accessibility for a global audience.
-
Translation Services
Leverages existing translation services and tools.
-
Global Reach
Supports diverse user base and enhances global reach.
-
Locale Alignment
Aligns with locales requiring native-language documentation.
-
-
Cons
-
Maintenance Overhead
Higher maintenance overhead and risk of inconsistencies.
-
Slower Cycles
Slower review and publication cycles due to translation delays.
-
Increased Costs
Increased costs for translation and localization services.
-
UX Fragmentation
May lead to fragmentation of the user experience across different languages.
-
Quality Challenges
Challenges in ensuring quality and accuracy across all languages.
-
Content Misalignment
Risk of misalignment between original and translated content.
-
Legal Risks
Potential legal and compliance risks if translations are inaccurate or misleading.
-
4.3. AI Agents¶
AI tools to assist with translation and quality assurance (QA) based on English.
-
Pros
-
Advanced Technology
Leverages cutting-edge technology for translation and quality assurance.
-
Cost and Speed
Potentially reduces costs and speeds up the translation process.
-
Reduced Human Error
Reduces human error in translation and QA.
-
Enhanced Maintainability
Enhances maintainability with AI-assisted updates and reviews.
-
Improved Accessibility
Improves accessibility with AI-generated translations.
-
Rapid Iteration
Enables rapid iteration and updates with AI support.
-
Multilingual Support
Facilitates multilingual support without extensive human resources.
-
Scalability
Provides scalability for handling large volumes of documentation.
-
Continuous Improvement
Allows for continuous improvement through machine learning and feedback loops.
-
-
Cons
-
Accuracy Concerns
AI translations may lack accuracy and nuance, leading to potential misinterpretations.
-
Investment Requirements
Requires investment in AI technology and training.
-
AI Dependency
Dependence on AI tools may introduce new risks and uncertainties.
-
Ongoing Evaluation
Ongoing evaluation of AI tools and their impact on documentation quality.
-
Security and Privacy
Potential security and privacy concerns with AI-generated content.
-
5. Consequences¶
-
Positive
-
Streamlined Processes
Streamlined documentation processes with reduced complexity.
-
Faster Publication
Faster publication cycles and improved responsiveness to changes.
-
Reduced Costs
Lower costs associated with translation and maintenance.
-
Consistency
Clear and consistent documentation in a single language.
-
-
Negative
-
Reduced Accessibility
Reduced accessibility for non-English speaking audiences.
-
Support Channels
Potential need for additional support channels for non-English speakers.
-
Market Risk
Risk of alienating non-English speaking users or markets.
-
6. Implementation¶
-
Update Documentation Standards
Update the contribution guidelines and documentation standards to reflect the single-language (English) strategy, including authoring conventions and review expectations.
-
Communicate to Stakeholders
Communicate the language strategy to all teams and stakeholders through onboarding materials and internal announcements.
-
Establish Review Process
Define a review process that ensures all documentation is authored, reviewed, and published in English before release.
-
Monitor and Evaluate
Periodically evaluate the strategy against evolving AI translation capabilities and compliance requirements, adjusting the approach as enterprise-grade solutions mature.
7. References¶
- Sentenz Docs as Code article.