Language evaluation principles and framework


The language evaluation framework is a guidance document developed by the Kubernetes Naming Working Group . The version of this documentation found on the Inclusive Naming website is a mirror.

It outlines a structured framework for evaluating language and terminology for harm to the community. This enables the community to navigate divisive conversations with a measure of clarity.

The framework was originally created for an open source technology project and was subsequently revised by the INI to apply to other technology-related contexts as well.

See also the implementation path document for a discussion of how changes in different portions of your project have different levels of impact on your users.

How to use this document

The framework is divided into three sections: first-, second-, and third-order concerns, ranked in order of potential harm to the community.

  • First-order concerns are language where harm is egregious, overt, and clearly problematic.
  • Second-order concerns are language which is problematic but with a less definite impact.
  • Third-order concerns indicate language that could use improvement but does no demonstrable harm.

Answer all questions for each term evaluated.

When complete, consider questions answered in the affirmative: in general, the more questions answered “yes” or “possibly”, the more likely it is that the language in question needs to be replaced.

If any first-order concerns are a “yes”, replace the language.

If a significant number of second- or third- order concerns are a “yes”, strongly consider replacing the language.

This framework is intentionally non-prescriptive. The intention in this work is to reduce harm for the community; let harm reduction guide your decisions.

See also the implementation path document for a discussion of how changes in different portions of your project have different levels of impact on your users.

First-order concerns

  • The term perpetuates bias or stereotypes in a manner that is harmful to any historically underrepresented group of people/
  • The term is overtly disparaging towards any historically underrepresented group of people.
  • The term is loaded, problematic or politically charged outside of technology contexts, even if the language is itself etymologically neutral.

Questions to ask

  • Does the term refer to a group of people in contexts outside of technology?
  • Is the term potentially disparaging in other contexts?
  • In discussion of the term, are opinions strongly split, even if the term isn’t directly harmful?

Second-order concerns

  • The term is unnecessarily violent, oppressive, or elitist terms.
  • The term originates from a history of injustice, oppression, or identity-based discrimination except when referring to the actual historical events.
  • The term is culturally appropriated
  • The term is associated with identity or status of a historically underrepresented group in a way that is negative or judgmental.
  • The term is unnecessarily gendered.

Questions to ask

  • Does the term have militaristic connotations?
  • Does the term or phrase have historical use either inside or outside technology contexts that someone can easily identity? - Did it arise during a world war, for example?
  • Is the term classist or otherwise segregational?
  • Is the term strongly associated with one gender?

Third-order concerns

  • The term is unclear, uses metaphor, or a more precise word is available.
  • The term anthropomorphizes (ascribes human traits to nonhuman objects) code or computer systems
  • The language is idomatic, and understanding is limited to those in a particular culture or in-group

Questions to ask

  • Does the term use metaphor, simile or other literary devices to communicate a point?
  • Is understanding the term conditional on belonging to a specific culture or in-group?
  • Does the term assign human-like qualities to technological systems?


This document is based in part on the framework for evaluating harmful language developed by the Kubernetes Naming Working Group, which is licensed under the Apache License version 2.0, and the Inclusive Prism Inclusive Language Principles, which is licensed under the Creative Commons-Attribution 4.0 International License.

Revision History

  • Version 1.0: Published March 2021 (ref: 3c059d1 )
  • Version 2.0: Published July 2021 (current)