Extended definition
Open Access (OA) is the academic publishing model in which content is free and openly accessible to readers, with no subscription or per-article paywall. The canonical definition comes from the Budapest Open Access Initiative (2002), Bethesda Statement (2003), and Berlin Declaration (2003), which articulated two requirements: free online access and reuse rights compatible with open licenses (typically Creative Commons). Suber (2012, MIT Press) is the canonical conceptual reference. Four operational variants coexist: gold (final article published in an OA journal, typically funded by APC paid by author or institution), green (deposit of preprint or postprint version in institutional/thematic repository with possible embargo), diamond (OA free for both author and reader, no APC, funded by academic societies or universities), and hybrid (subscription journal offering optional article-level OA via APC). Laakso et al. (2011) documented the growth trajectory since 1993, showing two decades of accelerated expansion.
When it applies
OA is appropriate in any project whose funding or institutional policy requires open access: research funded by ERC, NIH, Wellcome Trust, FAPESP (which has adopted OA policy since 2008), and most national funders with mandates today. It is a structural element of Plan S and similar mandates. Strategically, OA publication increases average visibility and citations (effect documented in multiple studies) and enables access by researchers in institutions with limited subscription budgets (especially in the Global South). For theses, dissertations, and technical reports, deposit in institutional repositories is the standard green-OA route.
When it does not apply
It does not apply in research with contractual confidentiality clauses, intellectual property embargo before formal protection, or data involving non-anonymizable participant privacy. It does not apply in manuscripts containing third-party material under restrictive copyright (figures, long quotes, proprietary data) without explicit authorization for open licensing. In some humanities and small disciplines, legitimate OA journals can be scarce and APC-based gold OA financially unviable — green deposit is the alternative. It does not replace peer review: many OA preprints have not been reviewed, and readers must distinguish status.
Applications by field
— Health and biomedical sciences: PubMed Central requires deposit of NIH-funded research; OA is the effective norm. — Physics and mathematics: arXiv and Open Research Europe consolidate green OA; many top journals are also gold OA. — Social sciences: SocArXiv, SSRN, and growth of gold OA in impact journals; Plan S pushed adoption. — Humanities: slower transition; OA books via OAPEN, Knowledge Unlatched still consolidating.
Common pitfalls
The first pitfall is confusing gold OA with mandatory APC — diamond OA is free for both author and reader and exists in thousands of legitimate journals (especially in SciELO, Redalyc, AmeliCA). The second is assuming OA guarantees quality — predatory journals use OA as a façade; verifying Scopus/WoS/DOAJ indexing and COPE adherence is a minimum practice. The third is confusing CC-BY with public domain — CC licenses preserve mandatory attribution and prevent uses contrary to the chosen license. The fourth is depositing the wrong version in green: many journals allow postprint but not the publisher’s final version; consulting Sherpa Romeo is standard practice. The fifth is treating APC as a quality signal — there is no consistent correlation between APC value and editorial prestige; high APCs reflect in part the market power of large publishers.