magic-linguistic-ethics

When to Use

• Any new dataset being considered for training/eval (early — before download).
• Endangered or Indigenous language data of any kind.
• Religious / sacred text use (Bible-NLP, Quranic, Vedic, Indigenous oral histories).
• License audit before release (open / community-gated / restricted decision).
• Attribution and provenance tracking design.
• Drafting a model card's "Ethics" / "Limitations" / "Intended Use" sections.
• Routing decisions involving community-controlled archives (DELAMAN, ELAR, AILLA, PARADISEC).

When NOT to use: the dataset is your own English-only synthetic data with no community attribution issues; or the operation is a pure technical refactor with no data implications. Even then, ask once — under-using ethics is the modal failure mode.

Why this skill is A-tier (and not optional)

A "good" engineer can build a tokenizer, mine bitext, fine-tune a model. None of that protects against:

• Training on a dataset whose community didn't consent to model use, even if the license technically permits it.
• Releasing a model that generates Sacred Indigenous content without permission.
• Stripping attribution lineage during a dataset merge so credit becomes irrecoverable.
• Using "open" Bible-NLP for a commercial chatbot whose generations the source community would not endorse.