User:SMcCandlish/Avoid quoting definitions

Directly quoting one particular reliable source (or ostensibly reliable one) on how to define any term of art (in the sciences, law, philosophy, the arts, a technical field, or whatever) term is something we almost always avoid, because attempts to define such things in monographs, papers, and topical dictionaries and encyclopedias virtually never coincide in their wording (unless directly quoting in their own right). A very common form of PoV-pushing on WP is the attempt to insert one particular quoted definition into an article, especially the lead, and it's a undue weight problem as well as PoV, often a sign of advocacy / soapboxing / "great wrongs" activism. It's an attempt to artificially narrow the scope and wording of the article to whatever suits one editor and their cherry-picked source; this is a form of original research.

Standard operating procedure

The normal WP practice is to paraphrase various cited sources on the meaning, and use a combined summary definition in the lead (within the limits of WP:SYNTH), citing in situ all of the ones used in the lead if that's the end of it; optionally going into the differences between and details of the definitions in later material (often moving the citations out of the lead to where they are more pertinent and less cluttering); and in cases of high levels of confusion or dispute, devoting a section to directly quoting and comparing definitions (sometimes there are even secondary sources we can cite that also do such comparison), either moving the citations to this section, or (when there's a lot of dispute) retaining them in the lead as well.

There's no rule and no clear need for one in this regard, it's just the way Wikipedian consensus has evolved through common sense. And we avoid a focus on definitions, rather than proper encyclopedic presentation, for an unmistakable policy reason. Two editors fighting on a talk page do not equate to a lot of dispute; when real dispute exists it will usually also exist in external publications. If it doesn't, this is a warning sign that someone may be PoV-pushing and that the "controversy" or "confusion" are being manufactured in a WP:POINTy way.

Pretense that basic dictionaries are good sources in technical contexts

General-audience dictionaries and encyclopedias are not reliable sources on their own for the field-specific meanings of terms; they are low-quality corroborating sources for the basic aspects of the technical meaning in some case, and high-quality sources for the general public's interpretation (e.g., how a journalist is likely to be [mis?]using a jargon term), but they all gloss over details that may be crucial in an academic context. They are often out-of-date when it comes to current usage; several fields may have developed a special senses of a term that are not found in any regular dictionary. Old dictionaries are not even reliable for general-audience usage.

Over-reliance on topical but iffy sources

The topical reference works take pains to not copy each others' definitions for copyright reasons as well as distinctiveness-of-our-work reasons. Monograph authors take is as part of their role (as does WP) to paraphrase previously published material and make it more digestible, and they exercise authorial as well as editorial judgement in the process, often introducing novel interpretations that don't necessarily agree with consensus in the field in question (when a monograph author does this, they are in fact a primary source not a secondary one for their original re-definition, and we must attribute the claim to the author, and are limited in how we can use it). Such sources need to be considered in the aggregate.

Journal papers are written for an expert audience, so if they introduce a definition, it's usually a contextually limited or modified one, as part of the methodology description ("For purposes of this research, the term ..."). In humanities journals especially, definitions are often explicitly provided to redefine something, in introduction of a new approach to something. New definitions are one of the many cases in which journal papers are primary sources (the default category for material in such sources, aside from literature reviews and systematic reviews).

Recent university-level textbooks (tertiary sources) often provide reliable, field-specific but generalized, definitions of such terminology, and are likely sources to use, but may not reflect nuances specific to sub-fields. The most reliable sources for discipline-specific definitions would be recent systematic and literature reviews in peer-reviewed journals, but they often avoid getting into definitions, which may be contentious within the field in question; such material is consequently difficult to find. They also tend not to define things that professionals in the field learned at university.

Tertiary sourcing is problematic

There are (at least) four problems with tertiary sources like dictionaries and encyclopedias, including field-specific ones:

1. They date quickly – some information in any of them is already obsolete before the work is published.
2. They gloss over details, and do not cover – or sometimes recognize the existence of – usage, views, and facts that are specialized (or in a different specialty).
3. They squirrel away related information into separate micro-entries instead of keeping it together, making it easy to miss entries and misunderstand the found ones – they lack context.
4. Like Wikipedia itself, they are highly selective in what they include; unlike Wikipedia we have no idea what their inclusion criteria are and what biases may be affecting them.

Continuing to pile on summarative tertiary sources at a Wikipedia article is usually not useful past a certain point of clarification and contextualization. We're supposed to be writing a tertiary work (that, unlike most, cites its sources), based mostly on secondary sources. We are not here to cannibalize and regurgitate other tertiary works, which themselves are just regurgitating other sources they do not identify. That said, there's no problem using such sources, within WP:CORE and WP:AEIS limits, for the contextualization we need in the short term (e.g. at a stub article), until better sources are found.