scholars are scholarly products that that bring together specialized source material, tools and expertise to support inquiry in a specific research area (Palmer, 2004, 2005).

Reading
The act of reading is a highly ubiquitous information activity that has rarely been the direct object of study in information behavior research. Thus, surprisingly little is known about the variable and complex reading processes involved in research and scholarship. Aspects of scholarly reading have been reported as part of more general studies of document or e-journal use, with the primitives of scanning, assessing, and rereading emerging in the literature. When information is first encountered, it is scanned in some preliminary way, as when a scholar reviews bibliographic fields while searching the online catalog or segments of pages when flipping through of a volume in a library. Each source is assessed to determine its relevance to the information problem at hand or to a longer term information need, and these interactions differ based on the kind of source and the researcher’s intentions and mode of inquiry. Other reading processes come into play when information is read more thoroughly or kept and reread later or over time.

General reading patterns related to e-journal use have been systematically documented in longitudinal surveys conducted by Tenopir and colleagues, showing differences among disciplines and important changes over time (eg, Tenopir, 2003; Tenopir et al., 2003; Tenopir et al., 2005; Tenopir & King, 2008). Not surprisingly, e-journal use has become the norm in the sciences and mathematics, where the format has been widely available for some time and readily adopted. Strong levels of use have also been documented in business and economics, but history, education and the arts have made a slower transition, due at least in part to lower levels of e-journal availability in disciplines outside the sciences (Education for Change, 2002). Such e-resource trends are suggestive, but they are not direct measures of actual reading activities.

Scanning
Researchers often begin working with documents by scanning them prior to engaging in more thorough reading. This practice has always been common with print materials, and it is accelerating and becoming more dynamic in the digital environment. For example, studies have demonstrated that scientists and engineers tend to skim papers to identify key components, beginning with the abstract, then moving to section headings, lists, summary statements, definitions and illustrations (Schatz et al., 1999). This process has extended to digital documents, where search features make it easier to pinpoint segments for reading, such as descriptions of experimental techniques or application of theories (Bishop, 1999). In particular, more recent large-scale transaction log studies have suggested that scholars are making greater use of abstracts in full-text databases (Nicholas, Huntington, & Jamali, 2007). Analysis of ScienceDirect logs showed that social scientists conducted the highest proportion of abstract-only sessions (41%), followed by mathematicians (40%), computer scientists (35%) economists (33%), life scientists (13%), engineers (13%), and chemists (12%). Supplementary surveys of users clarified that, while abstracts were valued for quick access and downloading, they did not substitute for reading the full article. Other studies have further confirmed that scholars often begin with preliminary parts of a document and then skim the full-text before printing for later reading (Tenopir et al., 2005).

The results from recent studies could be interpreted as evidence that scholars are reading more than in the past. For example, the number of articles read by university medical faculty was over 30% higher in 2006 than in the mid-1990s. At the same time, reading time per article fell, with medical scholars averaging about 24 minutes per article (Tenopir, 2006). However, while scholars are spending less time with more papers, they are also increasingly working through information on the Web by rapidly scanning material, or “bouncing” from site to site, a practice particularly common in medicine and the life sciences (Nicholas et al., 2006). Together these patterns suggest that researchers are not reading more, but rather scanning, exploring and getting exposure to more sources. In fact, they may be practicing active reading avoidance, as they quickly navigate through more material, spending less and less time with each item, attempting to assess and exploit content with as little actual reading as possible (Palmer, 2007; Renear, 2006, 2007).

Assessing
As researchers scan through documents quickly, sources of interest are assessed to determine their relevance and utility. This process of assessing has been described in number of ways—as differentiating, comparing and sifting—in previous studies (Ellis, 1993; Ellis et al., 1993; Foster, 2004; Unsworth, 2000). Bishop (1999) distinguished five separate stages of assessment, which can be distilled into these terms: “orientation” to form an initial impression of a work, “overview” to identify important details, “directing attention” to pinpoint specific document characteristics to skim, “comprehension” to interpret content, and “triggering” to initiate further reading (p. 265). Assessing is not always done item by item. Scholars have also been shown to assess aggregations of materials, such as issues of a journal or groups of sources that assist in keeping up-to-date with developments in their field of expertise (Bronstein, 2007).

A number of studies have focused on factors that influence a scholar’s decision to obtain, read or otherwise use a particular source. Key criteria identified among studies include topicality, originality, perceived quality, timeliness, availability and peer review (Covi, 1999; Houghton, Steele, & Henty, 2004). In a unique longitudinal study of agricultural economists, these and additional features— orientation, depth, reading time and author and journal reputation—were important assessment factors in decisions in the eventual selection and use of materials from sets of search results (Wang & White, 1999). For historians working with archival materials, assessing has been found to involve understanding the context of a source and its relationships to other sources in a collection (Duff & Johnson, 2002). As would be expected, book reviews are an important source of information for decisions about book purchases, according to a study of scholars across the arts, social sciences and sciences (Hartley, 2006).

Rereading
Books, articles, notes and documents of all kinds may be reread once or many times to recall content, increase comprehension and to relate and integrate previous research into a work in progress (eg, Brockman et al., 2001; Palmer, 2005; Tenopir et al., 2005). By revisiting previously read material, scholars also build their baseline of information, identify gaps in their knowledge and develop new research directions (Foster, 2004). Rereading is one of the primary reasons that scholars build personal collections. For humanities scholars, rereading a work is a significant part of interpretation and analysis, and it may be a long-term undertaking over the course of a project, across multiple projects over time or over a career (Palmer & Neumann, 2002). In the sciences, astronomers were found to engage in more rereading activities than other scientists, “which may be due to the ease with which astronomers can retrieve older articles electronically, or they may re-read more because they reuse older articles more than other scientists” (Tenopir et al., 2005, p. 793).

Rereading is closely associated with writing. A study of interdisciplinary humanities scholars found that collected texts were used to “prime” for writing activities (Palmer & Neumann, 2002). Music scholars frequently reread sources throughout the writing-and-revision stages of a project (CD Brown, 2002), and literary critics reread their assembled primary and secondary research materials in order to develop writing strategies (Chu, 1999). On the other hand, for historians, considerable writing may be necessary during an initial reading, especially with primary sources that must be accessed in archives and special collections, compared to journals, which can be reread later at their “leisure” (Case, 1991, p. 74).

Writing
As with reading, writing is another core scholarly activity that has not been studied in depth as an information behavior. The act of assembling information in constructing new scholarly works is one writing primitive discussed in the literature. In addition, co-authoring and disseminating are distinct activities of interest in information research, with dissemination having a strong base of more general, relevant literature that has developed in the area of scholarly communication over a number of decades. Differences in writing structure and style are apparent across disciplines and reflect scholars’ approaches to formulating, articulating, organizing and presenting evidence within the research process (Cronin, 2003). “Writing is not just another aspect of what goes on in the disciplines, it is seen as producing them”; each act of writing reconstructs and reinforces existing practices in a given field (Hyland, 2000, p. 3). A few studies have reported on levels of information use associated with writing processes. Scholars working in literary criticism were found to continue extensive information use throughout writing stages of a research project, while information searching tapers off to moderate or low levels (Chu, 1999). In a study of neuroscientists, both searching and reading were shown to continue during the writing phases in experimental and informatics projects, particularly to judge how to discuss new findings and claims in relation to existing literature (Palmer, Cragin, & Hogan, 2004).

Assembling
During the research process, scholars compose their thoughts through writing. Parts of texts are composed, integrated, revised and refined as the foundation of new publications. Ellis (1993) used the term “assembly” to denote the myriad of processes that are involved in drawing together ideas and results and writing them up for publication. Assembling is practiced in concert with searching, as well as reading, and extends well into the writing phases of a research project. Assembly by interdisciplinary scholars has been described as “picture building”—a set of behaviors scholars perform as they map out “in their minds, and on paper, the disciplines and concepts relevant to achieving an interdisciplinary overview of the topic” (Foster, 2004, p. 234). The cognitive and physical work of assembling a text can produce more than an image or sketch, however. It establishes the base for the scholarly product that will ultimately be disseminated. For humanities scholars, assembling a text is formative and iterative work that involves continual information management, accretion, and refinement. For example, one study described how “each scholar had his or her own way of taking pieces of an idea or passages that were excised for editorial reasons and putting these into new files or documents to feed into new papers” (Brockman et al., 2001, p. 27; Palmer & Neumann, 2002). Literary scholars have been found to write in multiple stages, assembling their analysis into initial drafts followed by further writing and re-articulation before dissemination (Chu, 1999). Multiple, simultaneous processes were observed in the research projects of art historians, who gathered and organized materials as they analyzed sources and structured their own written arguments (Bakewell et al., 1988). Other structuring approaches can assist in assembly, such as with music scholars’ creation of outlines, tables, lists, and chronologies early in the organizing stage of writing a research paper (CD Brown, 2002). Such non-narrative components appear to be especially influential in the sciences, where tables, diagrams and illustrations are a fundamental part of research communication (Hartley, 2006). Needless to say, word processing has altered assembly practices by simplifying integration, revision and formatting, and digital production has allowed for inclusion of multimedia content in assembled “texts.”

Co-authoring
The practice of co-authorship has increased in recent decades, and its escalation in the sciences has been particularly controversial. Cronin and his colleagues have contributed a series of studies on authorship and other kinds of attribution, finding, for instance, that the incidence of co-authored articles in a selected psychology journal increased over 55% from the 1930s to the 1990s, likely due to more quantitative and experimental approaches in the field and the rise of modularized and discrete tasks in research teams. There was a striking increase in co-authorship in chemistry as well, rising from 44% to 99% over the course of the 20 th century. In contrast, for philosophers—who tend to work independently with abstract issues and theories rather than with empirically based subjects, trends and data—only two percent of articles were co-authored in the 20 th century (Cronin, Shaw, & La Barre, 2003, 2004).

The phenomenon of “hyperauthorship” has surfaced in recent years in fields such as high energy physics and biomedicine, where large distributed research projects are common and can produce, in extreme cases, papers assigning over one hundred authors (Biagioli, 2003; Cronin, 2001). However, many of the researchers listed as “authors” on these articles have played a role in the research but not participate in the actual writing of the paper (Cronin, 2001). Preparing a manuscript for publication has been shown to require effective collaboration, and as such, actual writing tends to be handled by a small group (Kim & Eklundh, 2001). For those working in interdisciplinary fields, co-authorship can lead to significant information problems. Spanner (2001) found that differences in vocabularies made it difficult for research partners in computer science and biology to understand each other’s contributions. Construction of a research report by collaborators required many hours on tasks including translating terminologies and negotiating sentence structure and overall format. Palmer (2001b) showed that interdisciplinary writing can be further complicated by the need to explain concepts to new audiences. As observed by Cronin (2001), even when researchers writing a joint paper share a common disciplinary background, scientific writing can result in a “pasteurized prose of collaboration.” (p. 561).

Disseminating
Trends in open access tend to dominate much of the literature on dissemination of scholarly work, but there are other aspects of the dissemination process that have been studied empirically and are closely linked to the practice of research. For example, dissemination includes the work of evaluating appropriate journals and presses for submission of a manuscript and the act of presenting a paper and fielding follow-up questions at a conference. One project may spawn a number of scholarly products that emphasize different facets of the research or need to be crafted for different scholarly audiences or in variant formats. Extensive reshaping and rewriting is often required by referees during the peer review process, and there may be moderate use of new information for these rewriting purposes (Chu, 1999). Generalizations about the prominence of the journal article for scientific dissemination are well supported by research studies. Journal publishing has been documented as the major mode of dissemination in many fields of social sciences as well, with decisions about where to publish influenced by the standing of the journal in the field, followed by distribution and speed of publication, and audience to which the journal is addressed (Francis, 2005). Professional meetings are also considered essential dissemination routes, as seen in geography where “published information is too late” (Borgman et al., 2005, p. 647). Surprisingly, a study of music scholars found that conference proceedings were the most frequent mode of dissemination, followed by journal articles and then monographs (CD Brown, 2002). While more commonly associated with the humanities, books have been shown to be regularly produced in a range of fields, including psychology, linguistics and sociobiology (Varghese & Abraham, 2004). Books are prevalent in the humanities, but perhaps not as optimal as might be expected. A study of history, English and anthropology faculty at large research universities determined that, although departments expect monograph publication prior to consideration for tenure, the majority of faculty did not consider book-length texts necessary for representing or disseminating their scholarship (Estabrook, 2003).

Books can be particularly effective for broader circulation of ideas, however. They are more accessible for some audiences, providing an important means for presenting scientific knowledge to the general public, and they are especially important for the transfer of information across disciplinary boundaries (Palmer & Cragin, 2008). Dissemination of research •ndings to outside fields is challenging, since prestigious academic journals tend to be single subject publications and articles from outside the narrow focus are often rejected (McNicol, 2003). Therefore, scholars have difficulty determining where to publish interdisciplinary works and whether or not “journals outside ones’ immediate field will count for tenure and promotion” (University of Minnesota Libraries, 2006, p. 22). Levels of e-publishing have been rising along with the escalation of e-journal use. Although the sciences are generally associated with early advances in e-publishing, there is evidence that economists and computer scientists have been more reliant on the Web for disseminating information than scientists (Barjak, 2006). In the early RLG reports, psychology and chemistry scholars indicated that they preferred the longer peer-review process for disseminating articles rather than relying on preprint sources (Gould & Handler, 1989; Gould & Pierce, 1991). Since that time, faculty have developed more informed and positive perceptions of open-access and alternative models for publishing, but some scholars still perceive e-publishing to be risky and less rigorously reviewed. Studies have found that senior faculty tend to be more comfortable sharing early stages of work in online venues and that Web presentation and self-archiving is increasing across fields. For example, chemical engineering faculty have been shown to consider digital alternatives highly viable, and some archaeologists are now willing to share field observations on open-access sites (Harley, Earl-Novell, Arter, Lawrence, & King, 2007). As discussed by Kling and McKim (2000), “scholarly societies play a major role in the shaping of communications forums within a field, both because they are typically major publishers within a field, and also because they articulate and disseminate research and publishing standards for a field” (p. 1312). They note that both the American Chemical Society and the American Psychological Association have had policies directing authors not to put publications on the Web at any stage of production. A survey examining scientists’ use of e-print archives for dissemination reported that they were used by a small number of psychology faculty and less so by chemists who indicated it was “against the policy of the publishers.” Nearly one-quarter of psychology scholars also cited publisher policies as a reason for non-use of e-print archives (Lawal, 2002). The dialogue surrounding open access and the American Psychological Association’s position on online distribution of scholarship has deepened over the years (Bullock, 2004; Brehm, 2007). Some disciplines have long relied on pre-print servers for disseminating research results, the most renowned case being arXiv.org for physics, math, and computer science. Recently, the Consolidated Appropriations Act (2008) in the United States mandated that any research conducted on behalf of the National Institutes of Health must be made freely accessible, and other funding agencies like the National Science Foundation have been strong proponents of openly accessible research. Motivated in part by the rising cost of serials and the Web’s influence on scholarship, many universities across the world are developing their own institutional repositories (IRs) to preserve and freely disseminate the work of their scholars. The use of IRs by faculty has been associated with self-archiving behavior (eg, Kim, 2007; Xia & Sun, 2007). But while one international survey of over 1,200 scholars showed that nearly half of the respondents engaged in self-arching behavior (Swan & Brown, 2005), deposit in IRs has been slow in general. A range of factors have been identified, including faculty not understanding potential benefits and continued preference for traditional peer review venues over open access alternatives (Bell, Foster, & Gibbons, 2005; Crow, 2002; Palmer, Teffeau, & Newton, 2008; Park & Qin, 2007). At the same time, librarians and other proponents stress that IRs, author- pay models, and other open access options are “viable alternatives to the problem of unsustainable journal costs” (Harley et al., 2007, p. 8).

Collaborating
Research collaborations can range from two to hundreds of participants, as suggested by the hyperauthorship trends discussed above. They range along a “continuum” of engagement, from basic consultation to fully integrated teamwork, and project management may be loosely coordinated or highly structured and closely administered (Hara, Solomon, Kim, & Sonnenwald, 2003). Information exchange is a key component of successful collaboration (Haythornthwaite, 2006), as is support for administrative coordination and data storage and sharing, which are not yet sufficiently available to academic researchers, especially outside the sciences (University of Minnesota Libraries, 2006). The primitives of coordinating, networking and consulting are discussed, but data practices are covered later as a cross-cutting activity involved in collecting and collaborating. It is worth noting that team based research is most commonly associated with the sciences, in part because there are clearer divisions of labor in scientific research than in the social sciences and humanities (Borgman, 2007). However, studies have not always indicated consistent differences along disciplinary lines. The RLG studies showed that collaboration was highly valued by history and literature scholars, as well as those in chemistry, engineering, and physics. At the same time, scholars in art history, philosophy, anthropology and psychology reported that they did not typically engage in collaborative work (Bakewell et al., 1988; Gould, 1988; Gould & Handler, 1989). However, results of co-authorship studies presented above indicated an escalation of collaboration in psychology.

Coordinating
Collaboration requires coordination of group work, which becomes more complicated as the number of institutional partners and the distance between them grows, as demonstrated in an analysis of projects conducted as part of an interdisciplinary program funded by the National Science Foundation (Cummings & Kiesler, 2005). Coordination problems experienced by scientists and engineers ranged from issues related to software differences across sites to difficulty with relatively simple tasks, such as scheduling meetings. Direct supervision was found to be the most effective coordination mechanism, and, as more institutions became involved in a project, workshops were effective for fostering joint efforts. Although employing more coordination mechanisms generally led to increased success, the study also found that large, multi-institutional projects tended to use fewer coordination procedures, suggesting that “the work arrangements that make these collaborations possible require a deliberate strategy for coordination” (Cummings & Kiesler, 2005, p. 717). McNicol (2003) likewise argued that clear leadership and coordination were vital to successful interdisciplinary work, and both formal and informal communication channels necessary for managing joint activity. Early in a research project, collaborators need to define project boundaries and agree on “the doability of problems” to be addressed (Hara et al., 2003, p. 22). However, decisions on what work needs to be done, who is responsible for execution, and other details are also influenced by the structural, organizational and technological context and must be articulated and refined over the course of a project (Corbin & Strauss, 1993). Studies have suggested that socio-technical infrastructure can compensate, to some degree, for lack of physical proximity. Situation awareness, such as information about who has worked on what when, can be mediated by technology through “contextual, task and process, and socio-emotional information” that facilitates collaboration across physical boundaries (Sonnenwald, Maglaughlin, & Whitton, 2004, p. 990). Innovations in collaborative technology appear to be having an impact on research production, as illustrated by one study of Australian researchers, which reported that approximately 60% of respondents felt new information environments and technologies had changed the way that they collaborated (Houghton et al., 2004).

Networking
Before a collaboration can begin, relationships with colleagues and associates need to be established; and for a collaboration to succeed, those relationships must be strengthened and maintained. Information technologies are making it easier for collaborators to communicate and work together and, of course, e-mail has been the lifeblood of communication among local and distant team members for many years (Walsh, Kucker, Maloney, & Gabbay, 2000). As would be expected, one study documented that collaborating scientists depended on Web-based communication more than those who work independently, and larger teams showed higher levels of use (Barjak, 2006). In an earlier comparative study, networked communication was associated with a dramatic increase in co-authored papers in math, a field known for independent scholarship. In addition, use of electronic mail, bulletin boards and listservs was found to be more common in mathematics and physics than in experimental biology or chemistry (Walsh & Bayma, 1996a, Walsh & Bayma, 1996b, p. 689).

Across one campus, humanities scholars and social scientists considered their colleagues to be everywhere, regardless of discipline, department, institution or even country (University of Minnesota Libraries, 2006). The traditional invisible college (Crane, 1972), or network for exchanging vital information in a research area, has become more of an “invisible constituency”—a heterogeneous, open and loosely organized network that serves more as ad hoc consultation than gatekeeping (Palmer, 2001b). New kinds of digital forums are now increasing engagement among researchers, resulting in online communities that foster collaborative research (Bibliographic Services Task Force, 2005). However, technology should not be considered the sole, or even primary, force producing these changes (Walsh & Bayma, 1996a, 1996b). Which technologies are chosen and how they are implemented is understood to be influenced by social and cultural factors specific to the research community (Hara et al., 2003), and therefore uneven adoption across scholarly communities is to be expected, as researchers gravitate to tools that fit the needs and practices of their collaborative groups.

Consulting
Scholars rely on consultation to assist with a number of scholarly activities. As discussed above, they contact colleagues and other experts for assistance in identifying information in the chaining process, and personal collections can function as a valuable consultative resource for other scholars. Researchers also regularly consult with each other to generate and test out ideas or to verify that they are following a productive and competitive research path. Consultation for stimulating and refining ideas has been observed in studies of scientific fields like molecular biology and neuroscience but also in music, history and the humanities more generally (CD Brown, 2002; Case, 1991; Covi, 1999; Palmer, 2005; Palmer et al., 2007). Historians who work primarily with archival materials consult about specific information sources, conferring with archivists as well as other researchers who have interacted with the materials being studied (Duff & Johnson, 2002; Palmer & Neumann, 2002).

In the networked information environment, scholars are more easily spanning geographic and intellectual space as they consult. The Web has been shown to foster short-term encounters with distant acquaintances or strangers that require little effort but have potentially high returns in access to valuable papers and bibliographies (Kuruppu & Gruber, 2006; Palmer, 2005). These associations are akin to what Cronin (2005) referred to as “cognitive partners,” or the “unwitting, occasionally unseen, and not infrequently sidelined helpers” that support the scholarship of others (p. 110). In the humanities, the high level of dependence on these consultative relationships can “approach joint authorship” in terms of influence on a publication (Brockman et al., 2001, p. 11). For interdisciplinary humanities scholars, consulting with important scholars in outside areas may be necessary for translating ideas from one disciplinary context to another (Palmer & Neumann, 2002).

Cross-cutting primitives
Thus far we have focused on primitives that occur as scholars conduct information work related to a particular activity: searching, collecting, reading, writing or collaborating. Some primitives, however, naturally straddle or cut across two or more information work activities. In this section, we discuss four such cross-cutting primitives. The first three—monitoring, notetaking and translating—are of interest because of their significance in the research process but, unfortunately, there is a limited amount of research from which to draw conclusions. The fourth category, data practices, stands out from the others. It is not a primitive in its own right, but a set of activities around which a growing body of discourse and new research is emerging. The literature does not yet lend itself to identifying discrete primitives. However, it is an area of vital importance, due to the current emphasis on e-science and cyber infrastructure in the information professions, and across fields where support for digital scholarship is a concern for researchers, universities, funders and others with interests in advancing the research enterprise.

Monitoring
Although directed searching, chaining, browsing and probing all play integral roles in scholarly information-seeking, it is also useful for scholars to review new, relevant information on a regular basis. This type of monitoring behavior was defined by Ellis (1993) as “maintaining awareness of developments in a field through the monitoring of particular sources” (p. 482). Another study building on Ellis’s work identified four types of monitoring differentiated by the type of source material (Bronstein, 2007). Monitoring electronic materials involves “performing a periodical literature search on abstracting and indexing databases, library catalogues, or Web sites to keep up-to-date with developments in the field.” Monitoring printed materials involves “periodically looking for new book reviews, or looking through new journals issues.” Networking consists of informal communication with colleagues in order to follow new developments in a research area.

Finally, citation tracking is a “consequence of the different monitoring activities,” and involves chaining activities or accessing other forms of referential materials to locate new sources. These forms of monitoring illustrate its cross-cutting nature: networking with colleagues and citation tracking, or chaining, are primitives associated with collaboration and searching activities, and reviewing journal issues may involve both browsing and scanning, primitives associated with searching and reading activities. Humanities scholars and scientists have consistently reported frustration trying to keep up with information in their fields and the need for monitoring activities to stay current (Borgman et al., 2005; Brockman et al., 2001; CD Brown, 2002; Murphy, 2003; Tenopir et al., 2005). Reading journals received through personal subscriptions has been a traditional strategy (CM Brown, 1999), but, as indicated above, the rate of personal subscriptions has been decreasing in recent years. Nonetheless, studies continue to report that scanning new issues of journals is the most common way scholars monitor developments in their field (eg, Francis, 2005; Vakkari & Talja, 2006), as newer Web-based services such as RSS feeds and citation alerting services are also being adopted. As early as 2000, the SuperJournal Project showed that the majority of users valued alerting features (Eason, Yu, & Harker, 2000). Newer studies have documented increased use by of listservs, RSS feeds and other automated services by scientists (Hemminger et al., 2007). More specifically, Tenopir et al. (2005) found that many astronomers search current awareness resources online, but fewer used services such as the Astrophysical Journal (ApJ) Yellow Pages or emailed tables of contents. A study of scholars at a Malaysian university showed a preference for e-mail alerts linked directly to articles or table of contents of a particular journal (Zainab et al., 2007). Although there is less evidence of wide-scale adoption of alerting services in the humanities, one study indicated that interdisciplinary humanities scholars favored “push services” such as subscriptions, listservs and mailings for keeping up with current trends in research (Palmer & Neumann, 2002). Personal contacts are important sources of information for monitoring. The RLG studies showed that physicists, astronomers, computer scientists, political scientists and anthropologists relied on electronic communication and in-person meetings with colleagues to keep up with research developments (Gould & Handler, 1989; Gould & Pierce, 1991). Other approaches applied by social scientists and humanities scholars included attending scholarly conferences and colloquia, consulting book reviews and scholarly association newsletters, and reviewing preprints or reports of research in progress (Bakewell et al., 1988; Gould, 1988; Gould & Handler, 1989; Gould & Pierce, 1991). Westbrook (2003) found that women’s studies scholars tracked the personal home pages of researchers known to them in the field. Currently, the online communities formed for collaborative purposes (discussed above) are a growing part of the repertoire of sources used in the monitoring strategies applied by scholars.

Notetaking
Discussions of writing practices have often focused on scholarly publishing, with little attention to how writing contributes throughout the scholarly production process. Above, we covered the primitive of assembling, in which notetaking is an important part of the writing done in preliminary stages of constructing a text. Notetaking is also a significant part of searching and reading, and in fact may be practiced together with any scholarly activity. Notes are produced systematically—on paper and online, in lab and field notebooks, and as part of data collection, experimentation and other more informal processes. Scientists record ideas, comments and procedures to accompany data, and scholars in all fields make annotations to articles they read and documents they write. Studies of annotation practices are informing the development of reading devices and writing software (eg, Marshall, 1998; Marshall & Bernheim Brush, 2004; Schilit, Golovchinsky, & Price, 1998) and tools for assisting scholars in documenting their work with digital libraries and other online content (eg, Bradley, 2008). In the Web environment, however, development of annotation systems needs to account for the fact that many individuals expect their notetaking to remain private and that there is a difference between “idealized memory” encoded in an annotation and its actual value (Marshall, 2005).

It has been noted that in the sciences, the “scribbling” and “jotting” of ideas and other informal writing that is performed at the bench may be a better representation of scientific work than the formal writing presented in research papers (Rheinberger, 2003). Notetaking is also widespread in the humanities, produced on all kinds of documents and in the course of managing all aspects of physical and intellectual scholarly work, from coordinating sources materials to generating original new texts (Brockman et al., 2001; Case, 1991; Toms & O’Brien, 2008). Studies have shown that literary scholars use mapping, sketching, and outlining for recording notes; historians develop elaborate personal notetaking systems that emphasize chronology; and music scholars systematically capture explicit musical examples to be used in their written works (CD Brown, 2002; Case, 1991; Chu, 1999). Notetaking done in tandem with other writing tasks can be a largely tacit process that can produce large amounts of structured but informal text (O’Hara, Taylor, Newman, and Sellen, 2002). As discussed above in relation to assembly, when scholars take notes, they “are not just documenting their ideas. The act of writing is formative” (Palmer & Neumann, 2002, p. 100).

Translating
Scholarly work that crosses disciplinary boundaries poses a unique set of challenges. For scholars who are classically trained in a discipline, navigating the literature and research practices of another field requires developing familiarity with new terminology, concepts, theories and methods. For interdisciplinary collaborative groups in the sciences, translating is part of learning about collaborators’ perspectives in relation to mutual research interests, and it is a necessary part of the communication required for making research progress (Palmer, 2001b). In the humanities, collaboration tends to be less formal in nature, but those involved in interdisciplinary scholarship still must translate as they work with sources and people outside their field and as they write for other disciplinary groups (Palmer & Neumann, 2002). In a study of environmental scientists, 89% of respondents indicated that they needed to be somewhat or very familiar with the terminology of another discipline in order to understand literature they were consulting (Murphy, 2003). Similarly, among a small sample of humanities and social science interdisciplinary scholars, the majority indicated that they needed to become familiar with the vocabularies of disciplines outside of their primary field in order to conduct successful research (Spanner, 2001). Additionally, literature written for other disciplinary audiences must be assessed for potential source bias (Meho & Tibbo, 2003).

Colleague networks are essential for making and maintaining the greater number of personal contacts needed to share and validate interdisciplinary information (Foster, 2004; Spanner, 2001). Interdisciplinary humanities scholars, in particular, may depend on local colleagues or outside experts for assistance interpreting ideas and written material encountered from other domains (Palmer & Neumann, 2002). The process of co-authorship is also complicated by the need for continual negotiations and decisions on what needs to be explained to different audiences, the work involved in refining and clarifying terminologies, and agreeing on acceptable reporting structures and formats for different fields (Palmer, 2001b; Spanner, 2001).

Articles, Library management, Research
cultural policies, discourse analysis, public library