What is the difference between a conference paper and a journal article?
How they are reviewed, how quickly they appear, and how much they count — and the answer varies enormously by discipline, which is the source of most cross-field misunderstanding.
Journal article. Submitted to a journal, peer reviewed — frequently through multiple rounds of revision — and published when accepted. The process commonly takes many months and sometimes years. The result is the version of record, archival and citable.
Conference paper. Submitted to a conference by a deadline, reviewed against that deadline, and presented at the event. Published in proceedings.
Why disciplines differ so sharply:
In computer science, top conferences are the primary venue. They are highly selective, rigorously peer reviewed, and carry more weight than most journals. This reflects a field where developments move faster than journal timescales allow, and it genuinely confuses assessors from other disciplines who assume conference papers are minor.
In most sciences and humanities, conference presentations are for work in progress — early results, feedback, visibility — and the journal article is the substantive output. An abstract-reviewed conference talk carries little formal weight.
In engineering, practice varies between subfields.
What conferences provide beyond publication: feedback while work can still change; visibility for early-career researchers; and — frequently the most valuable part — the informal conversations, which is where collaborations begin and where the field's actual state is discussed rather than its published state.
The forms of conference contribution: full papers, short papers, posters, and abstract-only presentations, each reviewed with different rigour and counting differently.
Whether a conference paper can be published as a journal article. Usually yes if substantially extended — commonly requiring a stated proportion of new material — and journals set explicit policies. Submitting the same work twice without disclosure is duplicate publication and is misconduct.
Predatory conferences are a real problem: events soliciting submissions by mass email, accepting nearly everything, charging high fees, and sometimes not taking place as described. The warning signs are unsolicited invitations, implausibly broad scope, rapid acceptance, and organisers with no traceable academic standing.