Avoid default assumptions about ability, gender, or age

What it is

A default-assumption audit checks whether product copy treats one ability, gender, age, or life circumstance as true of every user. It asks whether an assumption is relevant to the task, supported by evidence, and compatible with other ways of completing it. It does not maintain a permanent prohibited-word list detached from language and context. The same word may clearly describe a real interaction in one place and create a barrier elsewhere by casting one body, identity, or age experience as the “normal user.”

Why it happens

Defaults hide in forms of address, pronouns, role examples, action verbs, help, and error attribution. Copy that says only “click” when keyboard, voice, and touch are supported turns one input method into the task itself. Inferring gender or age from a name, job, or purchase can produce incorrect template language. A repeated single prototype also shapes product decisions: teams explain only the path of the imagined “typical user” and misclassify other needs as exceptions. An audit traces each identity or ability cue to task necessity, data provenance, and alternative wording, separating necessary facts from unsupported inference.

Studying it

Sample high-exposure and high-consequence journeys. Annotate group references, titles, pronouns, age cues, sensory language, and device actions, then ask: is the trait necessary to complete this task, and did it come from the user, system capability, or author imagination? Invite participants with relevant usage experience to complete the real task and explain their interpretation. Observe whether copy makes anyone misjudge applicability, miss an alternative operation, or feel incorrectly classified. Record the specific context and affected task; one participant's view does not become a fixed language rule for an entire group.

Where it stops holding

Inclusion does not erase facts about ability, gender, or age. Healthcare, statutory eligibility, safety, and chosen personalization may require them, but the interface should state purpose, scope, and source. Grammatical gender, honorifics, and community self-designations need validation in the particular language. Lexicalized expressions such as “see” or “hear” are not automatically exclusionary; the issue is whether one sensory mode becomes the only usable path. Representation, roles, and authorization in people imagery belong to visual-asset review; this leaf concerns assumptions in product-authored language.

Applying it

• Add an assumption record to review of high-exposure and high-risk copy: implicated trait, task necessity, evidence source, potentially affected audience, alternative wording, and owner.
• Prefer objects and outcomes over guessed identity—for example, name display settings by function. Where relevant, use “select,” “open,” or “activate,” or explicitly describe the input methods actually supported.
• Do not infer title, pronouns, age, or ability from a name, avatar, voice, occupation, or behavior. Personalize only from authorized user-provided data applicable to the current context.
• Regress copy against the real interaction and record misunderstanding, abandonment, and discovery of alternate paths. Turn findings into contextual rules with examples, not an expanding automatic blacklist.

Related

• Same group: T1.08.2 Avoid metaphors with exclusionary overtones · T1.08.3 Self-described identity outranks system classification
• Adjacent: S3.02.4 People imagery must consider representation and diversity · T1.07.2 The target reader decides acceptable complexity
• Search terms: default-assumption audit · inclusive language · ability gender age assumptions