WikiJournal Preprints/Nice state history, if you can get it: Exploring open access and digital object identifier (DOI) registration in current U.S. state history journals

Introduction
State history journals, often published by state historical societies or universities, are an important forum for the dissemination of local and regional history. The internet offers tremendous possibility for the dissemination of works published in these periodicals, so much so that work which does not exist freely and accessibly online "will simply cease to exist as far as anyone but specialists is concerned", a condition which Patrick Leary refers to as the "offline penumbra". Leary's coinage of this term highlights two interrelated questions relevant to modern state history publications: 1) who are these journals for, and 2) who has what levels of access to them?

Tension around these questions has seemingly been felt by publishers and scholars throughout most of history of the American historical journal and its predecessors. The lineage of the scholarly U.S. state history journal is generally considered to have begun with Massachusetts historian Jeremy Belknap; in 1791, he published the American Apollo, an historical newspaper geared towards a popular audience. Although its readership quickly fell off and the paper folded in late 1794, Belknap continued with his goal of preserving and documenting American history in the Collections of the Massachusetts Historical Society, bound volumes of news, notes, and new society acquisitions aimed towards fellow historical society members. The New-York Historical Society took up Belknap's goal and model with a broad focus on American history, putting forth its own volumes of Collections starting in 1811. The country's centennial celebrations in 1876 spurred successful calls for new local history projects as the study of history was increasingly recognized as a academic discipline by the end of the 19th century. By the early 20th century, state historical societies became increasingly professionalized, while their journals transitioned from irregular volumes of bound collections to regular annual, biannual, or quarterly periodicals. The 20th century saw the establishment and disestablishment of numerous state historical publications and the ebb and flow of popular versus scholarly focuses as particular editors favoring one approach or another assumed control of different states' journals.

This study responds primarily to the second research question and briefly to the first by exploring the open-access status of ongoing peer-reviewed graduate and professional U.S. statewide history journals. "Open access" describes a gradient of possible levels of access granted by publishers for readers of academic literature. At the most conservative end of the open-access spectrum, a publisher might retain copyright on an article but release it free of charge to readers several years after its initial paywalled publication. Conversely, some deeply nonrestrictive open access levels allow readers immediate access free of charge to articles and their public peer reviews, and also permit unlimited sharing, modification, and reuse. Open access exists as one strategy in the larger ecosystem of digital scholarship distribution; importantly, for a work to be considered open-access (or not be considered open-access) it must be available online in the first place. The present research also considers whether any state history journals (or any portions thereof) exist within the offline penumbra, and considers the implications of this.

This study examines the registration of digital object identifiers (DOIs) for state history journals. DOIs are unique persistent identifiers generated by publishers through several registry agencies worldwide which are used to link metadata to scholarly works (books, journal articles, chapters, datasets, etc.) and redirect readers to landing pages where the scholarly work can be accessed (either for free or behind a paywall). DOIs are the most widely used persistent identifiers for scholarly works today, and their consistent application can increase the discoverability of research outputs. Whether or not they are consistently registered for a particular journal is here considered alongside the level of open access and general online availability of these publications as a metric of their publishers' continued commitment to persistent online access. Potential benefits of DOI registration for state historical journals are also discussed.

Finally, this paper is intended to serve as a catalog of active state history journals as of the present time. The task of developing a comprehensive list of U.S. state historical journals has proven challenging for at least a century since Augustus Hunt Shearer wrote in 1918 that "the number of these periodicals and the historical activities calling for their publication is not generally known." This sentiment was shared by Walter Whitehall in 1961 when he described the task of compiling an accurate list of these publications as effectively impossible considering the short lifespan and narrow distribution of some state history journals. Nonetheless, this study aims to provide a comprehensive snapshot of the U.S. state history journals actively being published circa 2020 as a reference for future scholars exploring the corpus of American state historical journals and their contents.

Methods
The present study focused on current peer-reviewed graduate and professional U.S. statewide history journals. These were found with the aid of several sources: EBSCO's America: History and Life (AHL) database, the community-generated Humanities Journals Wiki, and the Harvard Guide to American History. A Google web search for the term [state] history journal was also conducted in case resources not appearing in the previous sources were available elsewhere. Primary data collection took place between March 2019 and April 2020.

JSTOR, Project MUSE, and individual journal websites were consulted for each journal's publication history and details, accessibility, and preferred style guide. When information from AHL disagreed with information found on official journal sites, the information from the journal sites was preferred. Journal titles were checked against the OCLC's WorldCat and the University of Barcelona's Information Matrix for the Analysis of Journals (MIAR) to confirm bibliographic details. When DOIs were available on JSTOR, Project MUSE, or a journal's official site, the date of the earliest and latest issue to which they were registered was noted. Each journal's title was also checked on Crossref for DOI registration.

Inclusion and exclusion criteria
While the study of state history increasingly casts its focus beyond simple geographic bounds and while many journals in the present study acknowledged that contributions covering surrounding regions were also acceptable for publication, journals with explicitly regional or greater scopes (e.g., Middle West Review, the Journal of Southern History, or the American Historical Review) were excluded from the study, as were journals produced by organizations whose scopes were smaller than the entire state (e.g., the Journal of San Diego History or the Journal of South Texas) or dealing with a particular subset of state history (e.g., Afro-Americans in New York Life and History). Not all state historical publications are subjected to peer review (e.g., Illinois Heritage or the Wisconsin Magazine of History); those that were not were excluded. Journals publishing primarily or exclusively undergraduate research (e.g., the Iowa Historical Review) were also excluded from the study. Some journals asked authors to include citations in submitted manuscripts but due to nature of their popular focus, removed the citations from the published articles (e.g., Colorado Heritage or Washington state's Columbia: The Magazine of Northwest History); these were also excluded from the study as the lack of visible citations prevents any form of automated or manual bibliometric analysis. Finally, journals which no longer appeared to be active as of 2019 (e.g., Idaho Yesterdays, the Journal of Illinois History, or the Maryland Historian) were omitted. Journals whose names changed once or more throughout their publication histories were considered to be the same work if there were not gaps of more than one year between their old and new names. Some early irregular serial publications, such as the Collections of the Kansas Historical Foundation or the Collections of the Minnesota Historical Society which predated those organizations' respective journals were also excluded from the lineages of those states' modern journals.

When considering levels of open access, this study used modified versions of the access-level definitions offered by Piwowar et al. (2018). Their definitions applied to individual articles within publications, whereas this study considers access trends for those journals as a whole. To this end, when access levels varied from article to article within a publication, this study notes the least restrictive level of access that was generally available and indicates that this level of access is impartial ("imp."). Unless otherwise noted, an "imp." in Table 1 indicates that a journal's articles which are not covered by the least restrictive access level described are covered by the next least restrictive level (so "Bronze [imp.]" indicates that some articles are at least at a bronze open access level, while the rest are closed-access). By "generally available", an access level must be more frequent than a sporadic open-access articles, but need not be the access level for the majority of articles in that publication. From least restrictive to most, the access definitions used are:
 * Gold: All articles in a journal are published online at the same time as an in-print publication, if applicable, with an open-access license permitting at least unlimited reuse.
 * Bronze: All articles in a journal are accessible online free of charge via an official source but either all rights are reserved or the license status of articles is unclear.
 * Closed: All articles in a journal are accessible online only behind a paywall, either the publisher's or via a database like JSTOR, Project MUSE, AHL, or others.
 * Shadow: All articles in a journal do not appear to have been made available online in any form, either for free or behind a paywall. This definition calls upon Leary's concept of the "offline penumbra".

Other access levels like diamond, green, black, and hybrid are explored elsewhere but were not found to be useful for the study at hand and are not included here.

In general, print works in the United States automatically lose copyright protection and rise in the the public domain 95 years after their first publication. Thus, as of 2021, works first published prior to January 1, 1926, are in the public domain. For the purpose of access-level categorization, only articles still in copyright are considered. If a closed-access journal offers its pre-1926 articles for free outside of its paywall, this will not affect its designation as a closed-access journal. This paper will also note instances in which public domain articles are kept behind a paywall in spite of the expiration of their copyright protections.

Results
A total of 48 journals were analyzed in the present study, representing all U.S. states except Colorado, Idaho, Washington, and Wisconsin (see Table 1). Two states, Massachusetts and Pennsylvania, have two state historical journals apiece that fulfill the criteria and were examined. Of these journals, 29 are published quarterly, one is published thrice annually, 15 are published twice annually, and three are published once per year. Most of the journals surveyed (42, 87.5%) began annual or more frequent publication in the 20th century; just five (10.42%) began publication in the 19th century (including The Annals of Iowa, begun 1863, the oldest journal among the collection) and only one (2.08%; New Jersey Studies, begun 2015) was launched this century. The preferred style guide for all journals included in this study was The Chicago Manual of Style, as is customary for publications in the field of history. However, the recommended Chicago edition to follow ranged from the fourteenth, published in 1993, to the most recent version (seventeenth), published 2017.

Access levels
Just one journal (2.08%), New Jersey Studies, fit the gold access-level definition (see Figure 1). New Jersey Studies succeeded New Jersey History which ceased publication in 2013, and is published by the New Jersey Historical Commission, Rutgers University, and Monmouth University. Articles within the journal are licensed under a Creative Commons Attribution-NonCommercial 4.0 International License and the journal itself is registered with the Directory of Open Access Journals, an online compendium of open-access peer-reviewed journals without content embargos.

There are two fully bronze-level journals (4.17%) out of the 48 considered: Maryland Historical Magazine and Minnesota History. The entire run of the Maryland Historical Magazine is available for free on the publisher's website, although a free registration is required in order to access the materials. Minnesota History simultaneously publishes new issues in print and online, both paywalled on JSTOR and for free on the Minnesota Historical Society's website without requiring any form of account registration. Neither journal's articles are shared under a Creative Commons or other free license.

Twenty-one journals (43.75%) are accessible at in impartial bronze level, with at least some free-to-access articles but others kept behind paywalls. Within this category, there is a great deal of diversity in terms of how many articles may be freely accessed. On the most permissive end, journals including The Annals of Iowa and Kansas History have made their entire runs available online behind a one-year rolling wall, meaning that articles are published online one year after their initial print publications. On the other end of the spectrum, journals like Nebraska History, the Journal of Mississippi History, and North Dakota History offer only sporadic "featured" articles among their catalogs free to readers online, with most other articles paywalled. For the most part, articles in journals in this category that were not bronze open-access were closed-access. However, in a few cases, some articles appeared to be shadow-access, completely unavailable online even behind paywalls. This was the case for the entire first volume of Rhode Island History as well as articles in Ohio History published between 2004 and 2006, before which point articles are available on the publisher's site and after which point articles are hosted on Project MUSE.

A further 21 journals (43.75%) can be classified as fully closed-access: their articles are available online but are entirely maintained behind paywalls. Almost all of the journals in this access class are hosted on JSTOR or AHL which may offer free access to all articles to institutional subscribers or (in JSTOR's case) may allow for the download of a limited number of articles with the creation of a free account.

Finally, three journals (6.25%) can be classified as impartial closed-access journals: the Alabama Review, Delaware History, and the Historical Journal of Massachusetts. The Alabama Review can be found from 2007 onward on Project MUSE and from 1963 onward on AHL but its pre-1963 run appears to be shadow-access. Coverage of Delaware History begins on AHL in mid-1963 as well, with earlier issues also falling into the shadow-access realm. The Historical Journal of Massachusetts's coverage begins in 1980 in AHL, with issues published throughout the 1970s unavailable online in any form.

Public domain access
Of the 20 journals that began publication before 1926, all but one (95%) made content from 1925 and earlier free to access for readers. Nebraska History was the only journal that did not, although like other articles in the journal, this was due to the journal's inconsistent online presence: articles from pre-1926 that were available online were free to read and access, but many pre-1926 articles (like many post-1926 articles) were not accessible online at all.

JSTOR appears to automatically make articles that have aged out of copyright protection each year available for free, with the exception of the North Carolina Historical Review, which remained behind a paywall on the site. However, articles from pre-1926 through 1967 were available for free on the publisher's site. Conversely, articles in California History from before 1926 are free on JSTOR but not on the University of California Press's website.

Persistent identifiers
Only three of the 48 journals surveyed (6.25%) had registered DOIs for all articles published throughout their runs: California History, The Annals of Iowa, and New Jersey Studies. California History began its DOI registration with JSTOR, with more recent articles registered via the University of California Press, which publishes the journal. DOIs for The Annals of Iowa and New Jersey Studies, meanwhile, are registered by the University of Iowa and the Rutgers University Libraries, respectively. Of these journals, New Jersey Studies offers gold-level access, The Annals of Iowa offers impartial bronze-level access, and California History is distributed via closed access.

Another 17 journals (35.42%) include a mix of articles that have been assigned DOIs and articles that have not. All 17 of these publications have been made available either via JSTOR or Project MUSE. JSTOR tends to be more likely than Project MUSE to host a journal's entire run, while Project MUSE tends to host a journal's run from the point it agrees to work with the database (e.g., to host all issues from 2007 onward for journals that began working with the project in 2007). DOI coverage on Project MUSE is therefore more complete than on JSTOR, although JSTOR hosts a far greater number of articles than Project MUSE. In all instances but one, journals with impartial DOI registration contiguously cover newer articles without retroactively covering older ones; that is, in most cases, journals have DOIs assigned for their newest articles, usually within the past decade and a half, but not any articles earlier than that. For example, Texas's Southwestern Historical Quarterly, published since 1897, has DOIs registered for all articles from 2006 to present but none for articles published 2005 and before. The only exception is the The Arkansas Historical Quarterly, which has DOIs assigned for issues from 1942 through 2006, but not for more recent volumes. Of the journals in this category, six feature an impartial bronze-level of access, ten are fully closed-access, and one (the Alabama Review) is impartially closed-access.

Twenty-eight journals (58.33%), the majority of those in this survey, had not assigned DOIs for any articles in their entire runs. Journals in this category may be available in AHL, on JSTOR or Project MUSE, via publishers' websites, or through some combination of any of the above. While EBSCO, AHL's publisher, has a DOI prefix of its own, it does not appear to actively register many DOIs for journals in its databases. In this class, two journals are available under full bronze-level access (the Maryland Historical Review and Minnesota History), 15 are impartially available under a bronze level of access, 11 are fully available online via closed-level access, and two (Delaware History and the Historical Journal of Massachusetts) were available via impartial closed access.

Aside from DOIs, two journals use the Handle System to various degrees. The Handle System, which DOIs are technically a specific implementation of, can also be harnessed in other ways and has been packaged with some institutional repository software. This may have led to the adoption of non-DOI Handles (HDLs) as persistent identifiers for journal articles hosted by some institutional repositories, as appears to be the case for articles in the Hawaiian Historical Society and University of Hawai'i Press's Hawaiian Journal of History, hosted on the University of Hawaii's institutional repository. The Proceedings of the South Carolina Historical Association are also linked to HDLs, though these resolve at the issue level, rather than the article level, meaning there is not a specific persistent identifier for individual articles, but for issues as a whole.

Discussion
Joseph Amato writes that "People of every place and time deserve a history." If it is the work of local historians to provide "the natural link between immediate experience and general history", then state history journals are, in theory, one of the best venues for the completion of this task. They offer space for the rigorous interpretations and discoveries of historians, both amateur and professional, as they engage with the histories of their cities, states, and regions. Through state history journals, researchers may bring their scholarship to the public sphere and offer it into a permanent record.

The promise of the state historical journal as the vessel for immediate and specific histories, however, remains largely unfulfilled in the present ecosystem of their digital distribution. All but three of the journals surveyed kept at least some of their articles behind paywalls, with many maintaining their entire digital archives off limits to those unwilling or unable to pay for per-article access. Of these three, one still required prospective readers to register before gaining access to back issues, a needless barrier that might prevent the casual reader researcher from engaging with their state's history.

Open access is not a panacea. However, when used alongside additional paywall-free, public-oriented digital history projects, the consistent assignment of DOIs, and requirements that authors include DOIs in their citations, open access may go a long way towards bolstering the persistence and impact of U.S. state history journals and the research they contain. It will also offer an opportunity for work to be distributed more readily by scholars and readers, furthering the potential not just for increased readership but also to protect ephemeral digital content from lapsing into the offline penumbra. For instance, while the publisher site for the journal Idaho Yesterdays appears to have disappeared, because its contents were published under a Creative Commons license, individual scholars may freely host their articles on their own websites or in institutional repositories without fear of violating the journal's copyright.

Access levels for state history journals may be particularly crucial considering the often geographically niche content explored by researchers of state history. Matching scholarly output to an appropriately focused journal is an important part of the submission process, with an increased likelihood of publication for manuscripts that align neatly with their journal's aims and scope. Whereas scholars in some fields (medicine, for instance) have no lack of journals to choose from, including many open-access options, researchers of state history are in many cases limited to one obvious best choice for their manuscripts: the appropriate state's sole history journal. Several states have more than one focal journal as discussed in the Methods section, though these unsurprisingly tend to be journals devoted to the histories of older and/or more populous states. The dilemma comes especially for historians seeking to offer their research as open-access when their focal states have only a single dedicated closed-access journal. Most closed-access state history publishers do not appear to offer a hybrid open-access option in which an author pays a publication fee upon the acceptance of their manuscript to make it open-access immediately. This requires researchers publishing about states in which there are no state history journals with any form of open access to look elsewhere, possibly to a more regionally or topically focused hybrid open-access journal which may be more likely to be offered by a commercial publisher with the infrastructure and financial incentive to offer the hybrid model.

The application of DOIs for articles in state history journals varies widely, ranging from complete DOI registration to complete lack of DOI assignment. This is, unfortunately, consistent with publications in other humanities disciplines, which tend to trail the natural, social, and especially medical sciences in their registration of DOIs. As persistent identifiers, DOIs are not only useful for scholars and casual readers jumping between an article and its sources, but their presence also points to a commitment from a journal's publisher that an article will be persistently available for the long term. The consistent and complete registration of DOIs for articles throughout a journal's run thus signals a willingness from its publisher(s) to maintain a digital distribution scheme even withstanding the journal changing hands from one publisher to another.

The latest edition of The Chicago Manual of Style recommends including DOIs in citations for at least the articles that are consulted online, where DOIs are available. The consistent registration of DOIs for articles, along with the requirement that authors include DOIs in their citations, will make tracking a given article's scholarly impact easier and moving from a cited statement directly to its source less time-consuming. Sherriff's citation analysis of master's-level history theses indicated that historians tend to cite older materials than scholars in other fields, pointing to the usefulness of DOI assignment for all articles across a journal's run, not just the most recent. Sherriff also found that cited articles within history were not concentrated within a few journals but were dispersed widely among many titles; it follows that it is nearly as important for state history journals to register DOIs to their articles as it is for "bigger", more nationally focused journals. Studies have indicated that depending on the medium (thesis or journal article), journal articles made up between 7–15% of the total sources cited in history manuscripts. Registering DOIs and including them in citations will simplify bibliometric analysis, lessen administrative burdens for researchers, and make it easier for publishers and authors to understand how their research is being applied and understood by others.

The good news is that almost without exception, journal articles that have lapsed into the public domain are not hidden behind paywalls. Special commendation here must be paid to The Annals of Iowa which, in addition to making all but its most recent year of articles free to read online and registering DOIs for its entire run, has indicated the copyright status of every article in its collection. This includes an acknowledgement that its articles published from 1926 through 1978 are free from known copyright restrictions (making them essentially public domain materials) as copyright laws during that period required certain administrative tasks to be completed in order to maintain copyright that the journal's publisher appeared not to have completed. The State Historical Society of Iowa's forthrightness about this situation should serve as a model for other publishers; careful examination may indeed reveal that copyright on other state history journal articles from that time period was also not renewed, meaning that more material may be free than is currently known (or acknowledged).

Conclusion
Whether intended by its publishers or not, a resource can only be of use to those who have access to it. In the case of most current U.S. history journals, those who have digital access are primarily those who are either already institutionally affiliated (scholars and other researchers) or are willing to spend either money to buy per-article access to paywalled articles or time seeking out manuscripts from libraries or the authors themselves. However, it appears that even in serving these populations, U.S. state history journals have room for improvement.

For historians who seek to publish in state history journals, the largely closed-access nature of the articles in most states limits their potential readership. The general trend of articles being published without DOIs then also limits scholars' abilities to track their own scholarly impact from a research output. For readers who may already have access to state history journal articles, the general lack of open-access papers may increase the amount of time it takes to retrieve information. This is especially true for shadow-access articles, which cannot be retrieved without an in-person trip to a library holding these materials, or engagement with a interlibrary loan/document delivery service. The lack of DOIs also makes moving between sources more difficult and time-consuming, with the need for additional web- and database-searching. Finally, for potential readers, the general trend of closed-access articles makes even finding a resource difficult, let alone accessing it.

Historical societies and other publishers should look towards the journals that have successfully made their entire runs free to read and have registered DOIs to all their articles as models upon which to build the infrastructure for the next stage of local and regional history in America. Publishers must consider that their output will be the domain of those who have access to it: will it be purely the domain of a small subset of dedicated, hardworking researchers, historians, and academics, or will it also be the domain of everyone and anyone with an interest in the history of the place they call home?

Acknowledgements
Thanks to Elizabeth Fetterolf for awfully useful comments on earlier drafts of this paper and Elsa Mäki for help accessing journals when I had no institutional affiliation. Some findings from this study were presented as part of 4 Your Information 12 at The Tank in August 2021.

Competing interests
The author is presently a member support contractor with Crossref, a DOI registration agency. The conceptualization, data collection, and authorship of the majority of this manuscript were completed before the author began work with Crossref, who are not involved in or supporting this research.