deletions | additions       diff --git a/textbf_Discussion_and_recommendations_textbf__.tex b/textbf_Discussion_and_recommendations_textbf__.tex  index 188229b..c112b22 100644  --- a/textbf_Discussion_and_recommendations_textbf__.tex  +++ b/textbf_Discussion_and_recommendations_textbf__.tex 
...
Please note, as a preliminary remark, that for an Open-Access-by-default scheme to be fruitful, a particular regime is required also for what concerns the deposit which precedes the Open Access: it should in fact be mandatory and non-waivable. However, as we are going to deal with deposit \textit{infra} (§ …), we do not deepen that theme by now.  OA policies implementing Open-Access-by-default show to be effective even when accompanied by an opt-out option: \textit{“[T]he experience at every school with a waiver option is that the waiver rate is low. At both Harvard and MIT it’s below 5 percent”}, state Shieber and Suber \cite{shieber2013good}. Even the recommendations formulated for the ten years from the Budapest Open Access Initiative – whose motto, by the way, significantly was \textit{“setting the default to open”} \cite{budapest2012ten} – left to policy-makers the choice between granting or not an OA waiver to faculty (Recommendation 1.1 \cite{budapest2012ten}: “\textit{[...] When publishers will not allow OA on the university’s preferred terms, we recommend either of two courses. [...] Or the policy may grant the institution a nonexclusive right to make future faculty research articles OA through the institutional repository (}with or without the option for faculty to waive \textit{this grant of rights for any given publication)”} (emphasis added).), as if it did not weigh on the goals of the Initiative. Again, a recent research enumerated the elements of Open Access policies mainly correlated with policy effectiveness \cite{swan2015sept}: “Open Access cannot be waived” was not present.  On the other hand, the European Commission – in drawing up the OA regime to be adopted in Horizon2020 – has shown to be not so favourable to according an OA waiver: the Guidelines \cite{guidelines2016} do not mention it; the General Model Grant Agreement (H2020 GMGA) \cite{gmga2015}  seems to be a little more supporting, but only through really specific exceptions not related to an author's indepenent decision – \textit{“the obligation to protect results in Article 27, the confidentiality obligations in Article 36, the security obligations in Article 37 or the obligations to protect personal data in Article 39”} (GMGA, art 29.1) \cite{gmga2015}  –, and through very generic statements about author's will, like \textit{“}Unless it goes against their legitimate interests\textit{, each beneficiary must — as soon as possible — ‘disseminate’ its results by disclosing them to the public”} (GMGA, (\cite{gmga2015},  art 29.1 (emphasis added)). Anyway, note that the negative correlation between “Open Access cannot be waived” element and deposit rate has been stated (Report on policy.recording, p. 9). In fact, omitting a waiver option – making by consequence an author unable to publish with the journals of her choice \cite{shieber2013good} when they implement publishing rules inconsistent with the OA policy applied by the university – may push faculty to avoid deposit: \textit{“If authors have to worry about rights when making the decision whether to deposit in the first place, the cognitive load may well lead them to just not deposit”} (THE IMPORTANCE OF DARK DEPOSIT). As OA policy provisions requiring deposit – though strongly recommended for their effectiveness in promoting OA (\textit{infra}, §) – are not easy to materially enforce with regard to restive faculty (\textit{infra}, §), clauses discouraging deposit should be avoided.  Open-Access-by-default can therefore be partitioned in waivable OA and mandatory OA. Pay attention: sometimes a policy is described as “mandatory” or as a “mandate” even when it envisage an opt-out option. It may happen for two main reasons: i. the “mandatory” character actually refers to deposit, not to OA making (cf., e.g., SWAN); ii. as OA is set as the default, it is “mandatory” until the single faculty engage in active conduct to obtain waiver for a specific contribution (cf. PRIEST, p. 397). 
...
For the Berlin Declaration, deposit is one of the two conditions a contribution must respect in order to satisfy Open Access: \textit{“Open Access contributions must satisfy two conditions: […] 2. A complete version of the work […] is deposited […] in at least one online repository”} (Berlin Declaration).  The “Ten years on from the Budapest Open Access Initiative” states: \textit{“All university and funder OA policies should require deposit in a suitable OA repository between the date of acceptance and the date of publication”} (Recommendation 1.4 \cite{budapest2012ten}).  The European Commission, in its "Recommendation on access and preservation of scientific information" (2012/417/EU), does not go into details for what concerns the regime OA policies should apply (cf. in particular the section regarding “Open access to scientific publications”, §1), and therefore it does not mention what deposit scheme should be adopted. However, an indirect indication of the path Member States should follow – for what concerns deposit – in \textit{“[d]efin[ing] clear policies for the dissemination of and open access to scientific publications”} (2012/417/EU, §1) can be traced in Recital 11: \textit{“Preservation of scientific research results is in the public interest”}.  Moreover, in the context of Horizon2020, the European Commission requires deposit (GMGA, (\cite{gmga2015},  art. 29.2: \textit{“[Each beneficiary] must: (a) as soon as possible and at the latest on publication, deposit a machine-readable electronic copy of the published version or final peer-reviewed manuscript accepted for publication in a repository for scientific publications”}), which cannot be waived (In fact, the exceptions admitted by artt. 29.1 – \textit{“}Unless it goes against their legitimate interests\textit{, each beneficiary must – as soon as possible – ‘disseminate’ its results by disclosing them to the public”} (emphasis added) – and 29.2 – \textit{“This does not change the obligation to protect results in Article 27, the confidentiality obligations in Article 36, the security obligations in Article 37 or the obligations to protect personal data in Article 39, all of which still apply”} – of the General Model Grant Agreement only apply to \textit{“dissemination”} (i.e. to OA), not to deposit.). The European Commission, in its Guidelines on OA in Horizon 2020, moreover emphasizes that deposit must take place independently from the route the contribution will undertake for what concerns Open Access in the strict sense: \textit{“This step applies even where open access publishing ('gold' open access) is chosen to ensure that the article is preserved in the long term”} \cite{guidelines2016}. The “Decreto-legge 8 agosto 2013, n. 91, convertito, con modificazioni, dalla legge 7 ottobre 2012, n. 112”, art. 4, which has transposed the above-mentioned "Recommendation on access and preservation of scientific information" in the Italian judicial system, do not mention – precisely like the corresponding European text – the deposit regime to be adopted, but leaves to Italian public institutions the task to autonomously adopt the measures needed to promote Open Access (art. 4, comma 2: \textit{“I soggetti pubblici preposti all'erogazione o alla gestione dei finanziamenti della ricerca scientifica adottano, nella loro autonomia, le misure necessarie per la promozione dell'accesso aperto [...]”}). Therefore, given that a mandatory and non-waivable deposit actively contributes to OA policy effectiveness \cite{swan2015sept} \cite{shieber2013good}, such measures should include this particular regime for deposit.  If deposit is not waivable, that obviously means that there will be cases of “dark deposit” – i.e., of deposit of contributions which are not made OA. In fact, there can always be imagined cases of OA waiver: for example, when the policy allows it, either at faculty's discretion (we saw – \textit{supra}, § – that OA waiver is generally not opposed, and it is even recommended by some sources) or rather under particular circumstances (e.g., Italian policies always let an OA-adverse publisher triumph over them: supra, § [One may question whether the mere deposit too could succumb encountering a publisher opposing it. According to the AIB (Associazione Italiana Biblioteche), the answer should be negative, given that art. 67 of Italian law 633/1941 establishes a copyright exception for utilisation of works of authorship in the context of administrative procedures – among which we could also rank faculty evaluation –; as a consequence of this exception, any contract clause signed by the author which transfers to others rights on this type of utilisations does not apply (Osservazioni sul regolamento ANVUR). More generally, given that the mere deposit in the repository does not have independent economic significance, an author should be allowed to do it even if she transferred all her exploitation rights (GUIDELINES Parthenope, § 4). Again, the “Ten years on from the Budapest Open Access Initiative” – as you may recall (supra, §) – recommended dark deposit, and specifically \textit{“[w]hen publishers will not allow OA on the university’s preferred terms”} (Recommendation 1 \cite{budapest2012ten}): the eventuality of a publisher opposing dark deposit (and prevailing over it) was not even taken into consideration.]); or when OA can be avoided on a legal basis (e.g., when privacy or patentability are at stake). 
...
Luckily, the majority (nine) of the examined policies explicitly state that metadata are always OA from the moment of deposit. On the contrary, Universities of Bergamo, of Padova, of Trento, of Trieste, and of Udine do not specify that: therefore, from the plain text of this second group of policies it is not understandable whether metadata are immediately put OA, or rather they are unnecessarily submitted to the same procedures established for the deposited full texts.  As a final remark on the “deposit” matter, we should mention also the issue concerning the contribution version to be deposited.  Deposit of post-print versions (final peer-reviewed or published version) are strongly recommended (e.g.: FRANKEL-NESTOR, p. 15; GMGA, \cite{gmga2015},  art. 29.2; \cite{guidelines2016}, p. 5; Suber-shieber, p. 12). In fact, dissemination in OA of pre-print (i.e., not peer-reviewed) versions foments the belief that OA materials are of inferior quality, and therefore not reliable (cf., e.g., TEXT MINING OF SUBJECT: \textit{“in institutional repositories it is not always obvious what you will get back whereas searches in PubMed Central will only return peer-reviewed literature”}). Post-print can be distinguished in i. final peer-reviewed version and ii. published version. Generally speaking, we can state that they are both good. However, we consider important to underline that the European Commission’s Guidelines on OA in H2020 recommend to \textit{deposit the published version where possible [\cite{guidelines2016}, p. 5: “Where possible, the version deposited should be identical to the published version (in layout, pagination, etc.)”}. The GMGA per se does not manifest any preference (cf. GMGA, \cite{gmga2015}  art. 29.2).]. On the contrary, Suber and Shieber suggest to always deposit \textit{“the final version of the author’s peer-reviewed manuscript”} \cite{shieber2013good}, and to add to it – where publisher consents – also the published version \cite{shieber2013good}; they recommend however to not replace the former with the latter, but to store both, unless \textit{“the published version allows at least as many reuse rights as the author’s manuscript”} \cite{shieber2013good}. Obviously, post-print version should be required only with regard to those types of contributions which actually are submitted to peer-review (essentially, journal articles). Deposit of other content types which are per se not peer-reviewed should not be discouraged (cf, e.g., \cite{guidelines2016}, p. 5, which mention monographs, books, conference proceedings and grey literature; SUBER, which mentions ETDs).  However, in order to not threaten the trustworthiness of repositories (and of OA in general), the contribution type must be clearly noticeable and, if possible, repositories should confer users the possibility to filter contributions according to their being peer-reviewed or not (it might be useful, e.g., for purposes of text mining: cf. TEXT MINING OF SUBJECT).  One of the policy-makers we interviewed especially manifested her sorrow for the decrease in the deposit of such “grey literature” following to the shift from the previous institutional repository to IRIS (Institutional Research Information System): in her opinion, faculty perceive this System as much more visible and “official”, and therefore they refrain from deposit contributions other than journal articles. On the one hand, this faculty's reverence for the locus of deposit of their works for the purposes of OA may be useful in fostering a better quality of OA itself; on the other hand, it may deprive the Open Access patrimony of many worthy contributions. 
...
From its point of view, the “Ten years on from the Budapest Open Access Initiative” is less ambitious, as it doesn't require deposit of datasets, but anyway it recommends university policies to welcome repository deposit of them [Recommendation 1.1 \cite{budapest2012ten}: \textit{“When possible, university policies […] should welcome repository deposits even when not required (e.g. datasets […]).”}].  The OA policy recommended by the European Commission includes Open Access to research data: its Communication “Towards better access to scientific information” denounces how the publishing of such a resource has been neglected until that moment by the scientific research community [EC Communication Towards better access, p. 5: \textit{“Until now, scientific research results have been disseminated essentially by publishing articles. There is no well-established practice of publishing the underlying data.”}], and – on that same date – its “Recommendation on access and preservation of scientific information” consecrates its entire § 3 to \textit{“Open Access to research data”} (EC, 2012/417/EU).  In spite of this, art. 4 of the Italian “Decreto-legge 8 agosto 2013, n. 91, convertito, con modificazioni, dalla legge 7 ottobre 2012, n. 112” – which transposed the EC Recommendation – does not mention research data: it only refers to “articles”. However, as Guibault correctly underlined, \textit{“[t]he European Open Access Policy is not binding on the Member States, who are free to adopt the policy that best suits the needs of their own scientific community”} (GUIBAULT, Chapter 7, p. 400).  Nonetheless, European Open Access policy becomes binding in the context of Horizon2020: signing the General Model Grant Agreement [GMGA], \cite{gmga2015},  beneficiaries undergo the OA regime set by its article 29. And this article requires the deposit of \textit{“the research data needed to validate the results presented in the deposited scientific publications”} [art, 29.2 GMGA]: \cite{gmga2015}:  for example, \textit{“statistics, results of experiments, measurements, observations resulting from fieldwork, survey results, interview recordings and images”} \cite{guidelines2016}. However, it must be underlined that this requirement is merely an obligation of means, and not an absolute obligation: in fact, the beneficiary \textit{“must aim to”} deposit the research data [GMGA, [\cite{gmga2015},  art. 29.2; \cite{guidelines2016}, p. 5. This is confirmed by another sentence of the Guidelines, p. 5: \textit{“after depositing publications and, }where possible\textit{, underlying data [...]”} (emphasis added).]. Moreover, the OA making of research data is simply recommended, not required [\cite{guidelines2016}, p. 5: \textit{“Beneficiaries are also invited to grant open access to this data, but there is no obligation to do so”}.]. [By the way, this different OA regime applied to research data – in fact, deposit of scientific publications is on the contrary built as an absolute obligation, and their OA making is required – shows the inadequacy of ROARMAP's scheme of analysis under this point of view, as it merely refers to “Deposit of item” and “Making deposited item Open Access”, without distinguishing between the item “research data” and the item “scientific publications”.] We can therefore notice a certain discrepancy between the above-mentioned Recommendation by the European Commission (2012/417/UE), and the OA policy established by the same EU institution in the context of Horizon2020: in fact, the former asks Member States to \textit{“}[e]nsure that\textit{, as a result of these policies: research data that result from publicly funded research }become publicly accessible\textit{”} [2012/417/UE, § 3; emphasis added. Exceptions are however contemplated: \textit{“Concerns in particular in relation to privacy, trade secrets, national security, legitimate commercial interests and to intellectual property rights shall be duly taken into account"}.]. The regime established in the context of H2020 is thus less compelling than the one generally recommended to Member States. However, this discrepancy in the regime does not extend to the underlying orientation: indeed, an H2020 feature is its “Open Research Data Pilot” [cf. \cite{guidelines2016}, p. 7], and the General Model Grant Agreement includes an optional article (29.3) – i.e., an article beneficiaries may refrain from agreeing with while signing (except for, obviously, those beneficiaries leading projects covered by the scope of the Pilot) – requiring OA to research data. We can therefore notice that “required open access to research data” is still under assessment in the context of H2020; and this explains why its general regime for the moment simply recommends OA to research data. After all, the public consultation on Science 2.0 conducted by the European Commission in 2014 showed how some industry groups were against policy intervention in this area, one of them saying that \textit{“open access to data can discourage industry participation in research”} [Validation of the results of the public consultation on Science 2.0, p. 15, note 24]: a little of caution in introducing it is hence advisable; and Member States – in applying the EC's recommendation mentioned above (i.e., ensuring through OA policies public access to research data [2012/417/UE, § 3]) – should follow a line of action just as thought-out.  We already said (\textit{supra}, …) that, for the moment, the Italian legislator does not have imposed neither recommended deposit of and open access to research data: research data are not even mentioned by the national provisions transposing the “Recommendation on access and preservation of scientific information”. But Italian universities, what position did they take regarding research data? 
...
The orientation of the European Commission on this matter is not clear. Its “Recommendation on access and preservation of scientific information” [2012/417/EU, Recital 5] states that \textit{“[o]pen access policies aim to [...] enable the use and reuse of scientific research results. Such policies should be implemented taking into account the challenge of intellectual property rights”}: even if policies should aim to enable “reuse”, this term is not defined, and anyway IP rights must be taken into consideration in drawing the policies. In the context of Horizon 2020, we experience similar interpretation issues: here, derivative works are not mentioned in the list of re-uses – \textit{“the right to copy, distribute, search, link, crawl and mine”} \cite{guidelines2016} –; however, this list is not comprehensive, but only illustrative; moreover, authors are \textit{“encourage[d]”} – and not “compelled” – to choose CC-BY and CC0 licences \cite{guidelines2016}, just as in the “Ten years on from the Budapest Open Access Initiative”.  Regarding the second aspect, we can notice that the Budapest and the Berlin [cf. Also HOORN: \textit{“The first-mentioned condition [of the Berlin Declaration] can only be met if the rights holder decides that he or she wants to do two things, i.e. grant free access to end-users -which can be done in a straight-forward way - but also grant a licence to reuse the material for any responsible purpose”}.] Declarations, as well as the “Ten years on from the Budapest Open Access Initiative”, place on an equal footing free access and re-use: they are both required in order to realize Open Access. The same can be stated regarding the EC's “Recommendation on access and preservation of scientific information” [2012/417/EU, Recital 5].  The Open Access policy applied in the context of Horizon 2020 follows a midway path. The sole General Model Grant Agreement seems to require only free access for the purposes of Open Access: \textit{“Each beneficiary must ensure open access (free of charge, online access for any user) to all peer-reviewed scientific publications relating to its results”} [GMGA, art. 29.2]. \cite{gmga2015}.  However, interpreted on the basis of the related Guidelines: \textit{“[t]o meet this requirement, beneficiaries must, at the very least, ensure that any scientific peer reviewed publications can be read online, downloaded and printed. Since any further rights – such as the right to copy, distribute, search, link, crawl and mine – make publications more useful, beneficiaries should make every effort to provide as many of these options as possible”} \cite{guidelines2016}. Therefore, we can notice an absolute obligation for what concerns free access, and a relative obligation for what concerns libre access. Finally, the Italian “Decreto-legge 8 agosto 2013, n. 91, convertito, con modificazioni, dalla legge 7 ottobre 2012, n. 112” [“Decreto-legge 8 agosto 2013, n. 91, convertito, con modificazioni, dalla legge 7 ottobre 2012, n. 112”, art. 4, which has transposed the above-mentioned “Recommendation on access and preservation of scientific information” in the Italian judicial system.] only requires – for the purposes of OA – free access, and not also libre access [\textit{“[…] L'accesso aperto si realizza: a) tramite la pubblicazione da parte dell'editore, al momento della prima pubblicazione, in modo tale che l'articolo sia accessibile a titolo gratuito dal luogo e nel momento scelti individualmente”}.]. Notwithstanding the distance of the Italian provision from the international and European Recommendation it transposes – and from the other international and European OA texts – the Italian approach is anyway commendable, at least if we agree with the following assertion by Suber and Shieber: \textit{“Authors subject to this kind policy may use open licenses, such as Creative Commons licenses, to enhance user rights. The kind of policy we recommend here is compatible with the use of open licenses but does not require them. Institutions may adopt this kind of policy and decide afterwards when or whether to make use of open licenses. Similarly, it may adopt this kind of policy and leave authors free to make these decisions on their own, case by case”} \cite{shieber2013good}. According to them, it is therefore advisable to leave to authors the choice to allow re-use (pay attention: all type of re-use, not only the making of derivative works) when they make their contribution OA: OA can therefore be limited to free access. The above-mentioned Italian provision, by requiring from public institutions just those measures necessary to realize free access, precisely implement this suggestion, as it does not impose to institutional policies to require libre access. However, given that libre access is the most complete realization of Open Access, it is our opinion that the “Decreto-legge” should have at least mentioned it as something institutions should encourage.  After all, as the “Ten years on from the Budapest Open Access Initiative” correctly said: \textit{“In developing strategy and setting priorities, we recognize that gratis access is better than priced access, libre access is better than gratis access, and libre under CC-BY or the equivalent is better than libre under more restrictive open licenses. We should achieve what we can when we can. We should not delay achieving gratis in order to achieve libre, and we should not stop with gratis when we can achieve libre”} [Recommendation 2.1 \cite{budapest2012ten}].