Abstract
This paper reports the results of a study into a public space Internet portal which publishes guitar tabs (tablature) online, to examine what motivates people to participate in this activity and what benefits they get from doing so. A guitar tab is essentially sheet music for guitarists. The study examines why people contribute when it is easier for them not to publish their tabs and simply use the tabs that other people have posted. Answers to this will have implications for businesses wanting to encourage their employees to share their knowledge. An open ended questionnaire was sent to 183 tab publishers with a usable response rate of 39%, which is considered high for surveys. The questionnaire sought to gather data on motivations, benefits and community interaction. The paper begins with a review of relevant theories of knowledge sharing and publishing, in particular the private-collective model of innovation (von Hippel & von Krogh, 2003) which is used to analyze the results. Motivations are listed as under two categories, self and altruistic, with the most popular motivation being to share the songs with others, which is from the altruistic category. The most common benefit is personal satisfaction. The results show tab publishing fits with the private-collective model of innovation which means that a tab published online can be seen as a public good, as it is available to all, that has significant private elements. These private elements are the benefits that tab publishers get which the people who only use tabs without contributing their own, do not. The implications of the work are as follows. Enjoyment of the domain seems to be an important factor in motivating knowledge sharing. People who feel like they are part of a community and get satisfaction from being part of a community, will be more likely to contribute. The act of sharing knowledge should be as close to effortless as possible to encourage contributions. The act of preparing (collecting, collating etc.) the material to be shared should have meaning in itself for the person who is preparing it. If the act of sharing leads to increased status in the community people will be more likely to contribute. To encourage knowledge sharing, those who make use of the shared knowledge should be encouraged to give positive feedback to the person who shared it. To date, there has been little empirical work examining online posting forums.
Guitar Tabs and Knowledge Sharing
This paper reports the findings of a project examining why people contribute to a public space Internet portal which publishes guitar tabs (tablature) online. A guitar tab is essentially sheet music for guitarists, showing the notes and/or chords of a song without any information on the song's timing (for that guitarists must listen to the song). It is therefore easier to read and write than traditional staff notation sheet music (although musicians could debate this). The process of publishing guitar tabs involves listening to a song, figuring out by ear what the notes or chords are (transcribing the song) and documenting this in electronic form (tabs can easily be created using software such as Notepad). A sample tab is shown in Figure 1. The portal that was used in this study was OLGA, the Online Guitar Archive. The study examines why people contribute when it is easier for them not to publish their tabs and simply use the tabs that other people have posted.
An example guitar tab The lines are the strings, the numbers refer to the frets. This is the author's attempt at the introduction to I Can't Get No Satisfaction and is probably incorrect.
An example guitar tab The lines are the strings, the numbers refer to the frets. This is the author's attempt at the introduction to I Can't Get No Satisfaction and is probably incorrect.
A similar question has been asked in the literature about open source software projects. The open source model for sharing knowledge comes from the phenomena of highly talented computer programmers working together to write software which, when finished, will be available to anyone who wants it, for free. Often the programmers are contributing in their spare time and are not getting paid for their contribution. The best known example of a successful open source software project is the operating system Linux. Other projects include word processors and spreadsheet packages which have full functionality. Open source projects are typically started by a programmer (or a small group of programmers) who perceives a need for some software and asks for help writing it. This request is usually made on an open source Web page. When the project starts, this person will usually become the project manager, organizing the work so that there is no duplication of effort and ensuring that every module of software code works in its own right, and that it works together with everything else that has been contributed.
The rest of this section reviews work that has been done investigating the open source model of knowledge sharing, and an alternative model of knowledge sharing online, bulletin boards. The paper then discusses the method that was used to examine publishing guitar tabs and the results.
There are a number of things about the open source model that are difficult to explain. According to traditional economics it should not exist. Why should talented programmers work together in this way when they could be paid top salaries to write code? (In fact many open source programmers are also employed as programmers. Many others are students.) Why not simply wait for others to write the code, then use their finished software (so called free-riding)? It must be said that many people do use open source software without having contributed to it. Linux for example, has many thousands of users worldwide, and many of them could not have contributed to the code even if they wanted to, as writing software requires specialist technical knowledge. However this free-riding does not appear to dampen the enthusiasm of open source coders. On the contrary, they are keen to see their software get used. Actually, one way the general public are encouraged to support open source projects, is simply to use the software (Sourceforge.net, 2004).
von Hippel and von Krogh (2003) propose a private-collective model of innovation to explain contributions to open source software projects. They discuss two extremes of innovation, private investment and collective action. Private investment sees an innovator profiting from their work by intellectual property protection (copyright and licensing). Clearly, open source projects do not fit under this model as the innovator's intellectual property is given away. The collective action model on the other hand, is applied to the creation of public goods such as the construction of a public bridge or the opening of a public park. A public good is nonexcludable (Olsen, 1967). This means no one can be excluded from using it. It is this characteristic that leads to the free rider problem; even if users do not contribute to the public good, for example, by picking up some litter in the public park, they still benefit from a litter-free park because someone else will remove it. The city council get around this problem by charging tax and using that tax to pay someone to clean the park. However this cannot happen with an open source project. It has been suggested that free-riding can be overcome by applying peer-pressure to the group (Schwartz & Paul, 1992), something along the lines of a message saying “if you used this open source software, you should contribute to other open source projects,” although this does not happen. In addition, no effort is made to initially recruit programmers to get started on a project beyond posting a project idea on an open source website. So open source projects do not seem to fit perfectly under the collective action model either.
Since open source projects fit with neither model, the private-collective model sits in-between the two (von Hippel & von Krogh, 2003). This model believes open source software is neither a pure public good nor a pure private good, but is a public good that has significant private elements. It deals with the problems of explaining open source software in the following way. It proposes that free sharing of programming knowledge and ability does not represent a loss of profit for the contributor and that, in certain circumstances, there may be a net gain for contributors from network effects as the contributor's work gets used widely. Kollock (1999) discusses the possibility of producing an infinite number of perfect copies of knowledge, if it is electronic in nature. If it is, then according to Bergquist and Ljungberg (2001) the contributor can give away copies of their knowledge without diminishing its value. However, publishing knowledge can lose value for its owner (Boisot, 1998) even if it creates value for its user. For example if someone has a solution to a tricky problem others may be willing to pay for it. If the solution is freely available no one will pay for it and it loses value for its originator. This applies even when the value is not monetary. If someone has a recipe for some delicious dish they may be held in great esteem because of it. If more people can offer the same dish it losses its exclusivity and this esteem may be reduced.
The toleration of free riding is explained in a number of ways, primarily that programmers get benefits that users of the software, who have not contributed, do not get. These benefits might be feedback on, or corrections to, the contribution from other programmers; or agreements from programmers to contribute to other projects. Or the benefit might just be the satisfaction of creating something that is being used by others. von Hippel and von Krogh (2003) also point out that users of the software who cannot contribute to writing the code because they lack the skills required, can and do contribute to open source projects by reporting mistakes in the software. In this way, users become the project's testers. It is also recognised that in some cases software code is written for personal use and then, after the programmer realizes that it may have some use as part of an open source project, it is contributed. Such code might be a general purpose search algorithm for example.
Other writers have examined open source projects. Lerner and Tirole (2000) use economic theory to attempt to explain open source projects. They report that open source projects often stem from frustration with an imperfect existing solution that cannot be altered as the code is closed source (available only to the manufacturer). They also claim that giving credit to contributors is essential to keep them contributing and that increased reputation within the community does have a real effect on the programmers — many of them get jobs through their association with open source projects and there are cases where open source contributors have used their reputation to secure venture capital. They claim career concerns can explain much open source behavior. This fits in perfectly with the private-collective model — although knowledge is being given freely, the programmers involved get real benefits from doing so.
Bezroukov (1999), suggests a specialist academic model of participation for open source projects where programmers contribute, that is, publish their code, for reasons similar those that motivate academics to publish their research. Academics publish to receive peer review of their work, to contribute to the body of knowledge of their field, to get their ideas into circulation and, usually indirectly, for career benefits. Bezroukov suggests that programmers contribute to get feedback, which equates to peer review of their work. He also emphasizes the importance of a common enemy of the members of the community (in this case, Microsoft) in encouraging participation. Raymond (1998) has suggested that programmers simply contribute because they want to give a gift to the community. Mauss (1950), studying forms of exchange in archaic societies, sees the giving, receiving and refusing of a gift as a way to alter power relations. By receiving a gift, the receiver becomes subordinate to the giver, by refusing it the giver becomes subordinate.
Raymond and Bezroukov's views have been examined by Bergquist and Ljungberg (2001) in a study of gift culture in open source projects. They took the view that the contributions were gifts to the open source software community and ultimately users of the software, but agreed that the open source community is driven by similar norms as academic research. As such the gifts must be subject to inspection by the community for quality reasons, that is, they must be peer reviewed. Again this is explained by the private-collective model as the benefits to programmers may include feedback from peers. However, a key point to the private-collective model is that benefits can be derived from intrinsic or personal sources and do not have to be provided by the collective (the open source community). What this means is that a benefit such as satisfaction can be enough for someone to contribute; their contribution does not rely on someone else. These benefits are “private” and relate to the private investment model of innovation. Using tabs that have been published is a collective benefit, available to everyone and relates to the collective action model of innovation.
An alternative to the open source model of knowledge sharing is the bulletin board model. Bulletin boards represent a dialogue of questions from people requesting knowledge, and answers from people who possess that knowledge. Again the reasons why people post answers to requests have been discussed in the literature. McLure Wasko and Faraj (2000) studied three technical Usenet groups, which are a good example of the bulletin board model, splitting the reasons given for participation into three categories. The first was contributing to receive tangible returns. These might be in the form of access to expertise or getting answers to specific questions. People with reasons in this category gave to get. The second category found people participating to receive intangible returns such as a feeling of satisfaction or an enhanced understanding of the subject. The last category of reasons was that people contribute because they value the exchange of knowledge, interaction and feedback within a community.
In studying contribution to Usenet groups, Smith and Kollock (1996) use the work of Ostrom (1990) who studied a wide range of actual (as opposed to virtual) community activities. Ostrom indicates the importance of well defined group boundaries for communities to be successful, and having implicit or explicit rules governing behavior. These rules must be monitored for breakages. Smith and Kollock used Ostrom's work to examine why Usenet is a successful community. Usenet groups have clearly defined boundaries, the discussion topics, although these are not enforceable since anyone could post an off-topic message. Also, there are a set of rules that govern behavior. These are well publicized and are enforced by the group through flaming (sending multiple negative posts to offenders). Moreover, community members can filter out posts from repeat offenders. This effectively banishes them from the community from that member's point of view. Kollock (1999) lists future reciprocation, enhanced reputation and altruism as reasons for giving knowledge to online communities. Altruism includes helping because someone needs your help and giving help so as to feel part of the community. Rheingold (1994) suggests the reasons for contributing to online communities are to receive emotional support or intellectual companionship and having an interest in the domain of the community.
In fact the private-collective model can be used as a theory to explain the bulletin board model since again, contributions to boards do not fall easily into the private investment model of innovation or the collective action model of innovation; they have public and private characteristics. This time, the private benefits will be mostly intangible, the satisfaction of helping out someone in need for example.
Writing about knowledge sharing in general, Constant et al. (1994) investigated people's attitude toward sharing their knowledge finding that the more work experience or training people have, the more likely they will be to share knowledge, and that sharing expertise may produce personal benefits, so people are more inclined to share expertise than information. Jarvennpaa and Staples (2000) extend this work, considering the context of knowledge sharing such as the information culture of an organization. Bock and Kim (2002) investigated the motivators of knowledge sharing reporting that a positive attitude towards knowledge is a more important motivator than anticipation of a reward.
These two models for knowledge sharing, open source and the bulletin board, can be compared to determine which tab publishing is closest to. With the bulletin board model, knowledge is usually shared in response to a request; the work is not being published to get feedback on it. With the open source model, knowledge is shared because a programmer sees a need for additional functionality in the software and feels they can write the code to achieve it. So the programmer offers the knowledge because they think it will be useful, and not because someone has asked for it. This is similar to tab publishing. However, this distinction between the two is blurred. Sometimes organizers of open source projects will advertise the functionality that they need and programmers will publicly agree to write it. Likewise, with the bulletin board model, not all knowledge that is offered has been requested — a knowledge sharer may post a response that contains something along the lines of “I think you might also be interested in this….”
There are sites on the Internet that welcome requests for tabs, but OLGA is not one of them. Their attitude is very much if someone wants a song and it is not yet available that person should transcribe it and send it in. They welcome tabs with mistakes as this may encourage someone else to fix them. This appears closer to the open source model.
Another distinction between the two models is that contributors to open source projects will use the software when it is completed; that is, they will get benefit from the collective work of all the programmers. It is possible but perhaps less likely that those responding to requests on bulletin boards, who may be experts in the subject domain, will themselves gain new knowledge from the board. The only use a tab publisher will have for their own work at OLGA is as an aid to memory. They will however have use for other people's published tabs. So again tab publishing appears closer to the open source model.
Comparing writing open source code with writing tabs, writing code is much more demanding and takes much more time. Some coders claim then sometimes spend 40 hours a week writing code (Source forge.net, 2004).
Posting guitar tabs appears to be closer to the open source model of sharing knowledge but as has been explained this is not clear cut. With this in mind the private-collective model will be used as a theory to attempt to answer the research question, why do people publish their tabs?
Research Questions
Studies have shown that people often resist sharing their knowledge (Ciborra & Patriota, 1998) and yet OLGA is just one of a number of web portals giving access to thousands of songs. Even tabs for the most obscure songs can usually be found online somewhere. This research examines why the people who are involved in this activity share their knowledge with each other and with the rest of the world so freely. Specifically, the following questions are being asked, answers to which may have implications for businesses wishing to encourage their employees to share their knowledge throughout their organization: what motivates people to contribute and what benefits do they get (or want to get) from doing so?
Method
Whether or not publishing tabs is a legal activity has yet to be decided. Twice OLGA has been threatened by music publishers over copyright breach, and currently OLGA is trying to resolve the legality of by-ear transcriptions (OLGA, 2004). Certainly if lyrics are published there would be a legal issue. OLGA encourages people who publish tabs to “use lyrics sparingly” to show where chords changes take place (OLGA, 2002). At the top of each published tab is a disclaimer reading “This OLGA file is the author's own work and represents their interpretation of the song. You may only use this file for private study, scholarship, or research.”
Data was collected to gain insight into the activity of publishing tabs, specifically the motivations of the people who do this, the benefits they get from doing it, the interaction that goes on within this community and some personal details. It was felt that a questionnaire that could be emailed to a sample of the community would be the best data collection method as it would be quick to complete and easy to return, and by definition all potential respondents will be technically competent and able to cope with doing this. An open ended questionnaire was used so that respondents would not be constrained by a pre-determined list of motivations and benefits that they could select by ticking a box. It was important to get explanations from respondents, in their own words, of the reasons why they publish tabs. Appendix 1 shows the questions that were used. Several of the questions really ask the same thing, for example question 3 and 4, only the questions have been worded differently. This was done to really probe for the reasons involved and encourage respondents, some of whom might be quite young, to really think about the issues that are being asked and not just respond “because I want to.” It was felt that in order to get a reasonable response rate, the questionnaire would have to be quick to complete. For this reason, the number of questions was limited to ten, not including the personal details.
The questionnaire was piloted on 12 people chosen randomly from the sample of people who publish tabs. How the sample was derived is explained next. Five responses were received. Modifications were needed to make the questions more understandable to people whose first language is not English, as publishing guitar tabs is an international activity.
The email addresses of all participants were found from OLGA accessed from Harmony Central (http://www.harmony-central.com/Guitar/tab.html#static). The email address of each tab publisher can be found at the top of all of their tab files. Harmony Central organises guitar tabs by the song's artist. There is no published record of the names of the people who contribute tabs, so it is not possible to take a sample of them. Instead, artists were chosen at random; a random sample of people who publish tabs by that artist was then made. Unfortunately, the size of the active OLGA community is unknown and cannot be estimated. There is no list of contributors and if there was, OLGA is an archive going back ten years and many may no longer consider themselves as part of the community. 259 questionnaires were sent by email. 77 of the email addresses were incorrect (most likely the person has changed address) so the questionnaire did not reach its destination. Of the 183 that did, 75 were completed and returned. Of these, three were unusable. The usable response rate was therefore 39%, which is considered high for surveys. All responses were coded by the author. 20% of the responses were also coded by a third party, and Kappa was calculated for inter-rater reliability to be 0.88. Appendix 2 shows one of the responses and gives an explanation of how it was coded.
Results
The results are described and then discussed. Details about the respondents are shown in Table 1.1
Respondent information
| Age | Mean | 23 |
| Median | 21 | |
| Mode | 21 | |
| Sex | Male | 98.6% |
| Female | 1.4% (only 1 respondent was female) | |
| Occupation | Students | 46% |
| Other occupations were varied. 12% worked in some aspect of IT | ||
| Location | USA | 38% |
| Other locations were scattered around the globe. | ||
| Education | Most were well educated. Approximately 47% had or were working towards a degree. 2 |
| Age | Mean | 23 |
| Median | 21 | |
| Mode | 21 | |
| Sex | Male | 98.6% |
| Female | 1.4% (only 1 respondent was female) | |
| Occupation | Students | 46% |
| Other occupations were varied. 12% worked in some aspect of IT | ||
| Location | USA | 38% |
| Other locations were scattered around the globe. | ||
| Education | Most were well educated. Approximately 47% had or were working towards a degree. 2 |
Respondent information
| Age | Mean | 23 |
| Median | 21 | |
| Mode | 21 | |
| Sex | Male | 98.6% |
| Female | 1.4% (only 1 respondent was female) | |
| Occupation | Students | 46% |
| Other occupations were varied. 12% worked in some aspect of IT | ||
| Location | USA | 38% |
| Other locations were scattered around the globe. | ||
| Education | Most were well educated. Approximately 47% had or were working towards a degree. 2 |
| Age | Mean | 23 |
| Median | 21 | |
| Mode | 21 | |
| Sex | Male | 98.6% |
| Female | 1.4% (only 1 respondent was female) | |
| Occupation | Students | 46% |
| Other occupations were varied. 12% worked in some aspect of IT | ||
| Location | USA | 38% |
| Other locations were scattered around the globe. | ||
| Education | Most were well educated. Approximately 47% had or were working towards a degree. 2 |
Motivation
Motivations for publishing tabs online were grouped into two categories labelled self and altruistic. As questions were open ended, respondents were free to give more than one reason for publishing their tabs, so from the 72 responses, 130 reasons were given. These all fell into one of 11 different reasons. The self category includes reasons related to the person who published the tabs, the altruistic reasons were for the benefit of others. There was an exact split between the two; 50% of the 130 reasons fell into the self category, 50% into altruistic. The labels given to each of the reasons grouped in the self category are listed in Table 2. Also shown is the percentage that each appears in the list of 130 reasons.
Self category. Note figures are rounded to the nearest 0.5%
| Reason | Percentage |
|---|---|
| Ego | 13 |
| To improve guitar playing | 11.5 |
| Pleasure | 10 |
| To document tabs | 7 |
| To improve transcribing skills | 4 |
| To beat boredom | 2 |
| Return on investment | 1.5 |
| To avoid buying expensive music books | <1 |
| Reason | Percentage |
|---|---|
| Ego | 13 |
| To improve guitar playing | 11.5 |
| Pleasure | 10 |
| To document tabs | 7 |
| To improve transcribing skills | 4 |
| To beat boredom | 2 |
| Return on investment | 1.5 |
| To avoid buying expensive music books | <1 |
Self category. Note figures are rounded to the nearest 0.5%
| Reason | Percentage |
|---|---|
| Ego | 13 |
| To improve guitar playing | 11.5 |
| Pleasure | 10 |
| To document tabs | 7 |
| To improve transcribing skills | 4 |
| To beat boredom | 2 |
| Return on investment | 1.5 |
| To avoid buying expensive music books | <1 |
| Reason | Percentage |
|---|---|
| Ego | 13 |
| To improve guitar playing | 11.5 |
| Pleasure | 10 |
| To document tabs | 7 |
| To improve transcribing skills | 4 |
| To beat boredom | 2 |
| Return on investment | 1.5 |
| To avoid buying expensive music books | <1 |
Ego included any reason for publishing that included an element of desiring fame or increased status. Typical replies from respondents illustrating this reason are:
“get my name out into the world”
“for an ego boost”
“my name can be found on the search engines”
“to get my name “out there” on the web”
“seeing my name published online”
To improve guitar playing was listed by over 11% of respondents as a reason for publishing tabs.
Pleasure was taken to be any reason that included publishing tabs to get some sort of positive feeling. Responses here include:
“for my own pleasure”
“it's just nice to spread music around”
“I like playing guitar”
“greater sense of accomplishment”
“The reason why i do this is personal accomplishment. And maybe pleasure to know that I can be useful.”
The to document tabs reason means that songs are transcribed so that they can be stored in electronic form so the person does not have to transcribe them again in his or her head the next time they wish to play them. For someone giving only this reason, the publishing of the tabs is not the motivation for transcribing. Publishing is a secondary activity that comes after they are stored in electronic form. The following sums up this attitude: “[o]nce I've transcribed them for myself, it's not hard to send them to OLGA or another site.”
The reasons to improve transcribing skills and to beat boredom are self explanatory.
Return on investment is the label for giving to get; publishing guitar tabs with an expectation that at some point in the future the tabs of others may be used without guilt, or the feeling that requests may be legitimately made to others for tabs. Comments illustrative of this are:
“I hope someone will do something like that for me in the future”
“I hope that some time in the future they will feel the same and transcribe songs themselves”
Tabs are also published professionally in music books which typically cost as much or more than a music CD. To avoid buying expensive music books was listed by less than one percent (one person) as a reason for publishing tabs.
Table 3 lists the labels for the reasons falling into the altruistic category.
By far, the most popular reason given for publishing tabs online was to share the song with others. Some illustrative comments are:
“so others won't have to transcribe them”
“someone else might want to learn that song”
“perhaps there are other people like me who can't find the chords”
“help others”
“save others the trouble”
“I enjoy making people happy”
“I have worked them out and someone else might want to learn that song”
“for others to ENJOY ….cause i know that when i started playing and knew hardly nothing i really liked when i could find how to play songs i like that people transcribed, and it was really s****y not to find ones i love that no one tabbed ……that's why i did my site “the ideal tabs” for alot of unknown amazing bands i and a lot of other people around the world love and almost no one knows”
“To share my knowledge of guitar techniques with other guitarists, who maybe can't figure out the song I transcribe”
“To ease the frustration of a fellow musician in search”
Guilt is the label meaning that the person has used the tabs of others and feels they owe it to the community to publish their tabs.
To avoid others having to buy expensive music books suggests contributing to beat a common enemy, namely publishers who would profit from transcribing songs.
Benefits
When asked about the benefits received from publishing, 79 were given in total, which fell into one of 11 different benefits (which includes “none”). Table 4 lists the most popular ones along with the percentage each appeared in the list of 79. Each of the following appeared only once in the list:
Benefits from publishing tabs
| Benefit | Percentage |
|---|---|
| Satisfaction | 32 |
| None | 29 |
| (Positive) feedback | 22 |
| Fame in the community | 9 |
| Benefit | Percentage |
|---|---|
| Satisfaction | 32 |
| None | 29 |
| (Positive) feedback | 22 |
| Fame in the community | 9 |
Benefits from publishing tabs
| Benefit | Percentage |
|---|---|
| Satisfaction | 32 |
| None | 29 |
| (Positive) feedback | 22 |
| Fame in the community | 9 |
| Benefit | Percentage |
|---|---|
| Satisfaction | 32 |
| None | 29 |
| (Positive) feedback | 22 |
| Fame in the community | 9 |
“inspire others to play guitar”
“free albums from the group I transcribe for”
“intellectual stimulation”
“a permanent document of my songs”
“make friends from the community”
“improve guitar skills”
“improve computer skills”
The following two benefits were given tongue in cheek and are not included in the list of 79:
“make a better world”
“make people smile”
Interaction
Table 5 summarises the interaction that goes on between the people who publish tabs and each other, and the people who publish tabs and the people who use their tabs, who may or may not also be people who publish tabs — there is no way of knowing.
The questionnaire included a question about whether respondents keep a record of which tabs people are downloading (question 7). It was asked to give further insight about the interaction within the community. However there is in fact no way of doing this at OLGA so every response was negative. For question 9 about using other people's tabs, every respondent said they did use others' tabs to learn to play the songs. One claimed that he looked at others' tabs after he had transcribed the song himself, to compare them. Finally, the questionnaire asked how long on average it took to transcribe a song. Answers varied from less than a minute to a few months.
Discussion
The self motivation category equates to the private investment model of innovation, the altruistic category to the collective action model. It can be seen, as with open source projects, publishing tabs does not fit exactly under either model. As von Hippel and von Krogh (2003) point out, in traditional private investment of innovation, manufacturers produce innovations for users, whereas with publishing tabs it is the tab users who are producing the tabs. In addition, under the private investment model, intellectual property is licensed and sold. In the domain of guitar tabs this equates to publishing the tabs in a music book sold at a sheet music shop. Clearly this does not happen with tabs published online. These deviations from the private investment model can be explained, as has been found, by the tab publishers profiting (getting private benefits) in some way from publishing their tabs. These benefits were listed in the results section. So publishing tabs has some private investment elements.
However publishing tabs online makes them public goods. This brings with it nonexcludability and the free rider problem. The toleration of this can be explained in a number of ways. Firstly, tab publishers do not mind their work being used by others who do not publish tabs. In fact as has been shown, they get benefits from this (fame in the community and feedback on the work for example). Secondly, these benefits are unavailable to people who use tabs but do not publish them.
So tab publishing sits between the private investment model and the collective action model and fits well under von Hippel and von Krogh's (2003) private-collective model. Thus, tabs published online can be seen as public goods that have significant private elements.
von Hippel and von Krogh (2003) do not make a distinction in their work on open source projects between a motivation for writing open source code and a benefit from doing so. In this research project, the distinction between motivation and benefits is blurred. A psychologist may be able to give a precise definition of the two but it is expected that not all respondents knew exactly what the difference was (motivations are the factors that made the person publish and benefits are the positive things the person gets after having published). Each respondent could have listed as their motivation “to achieve the following benefits.” No attempt was made to explain the difference at the beginning of the questionnaire as it was felt that having to read and understand a paragraph of definitions might put people off responding. Likewise the distinction between some of the motivations is blurred, there are overlaps between them. Two possible examples of this are to improve guitar playing and pleasure - improving guitar playing (hopefully) is pleasurable and to share the song with others and guilt— the motivation to share the song with others may stem from a feeling of guilt. For this reason no in-depth statistical analysis on the results has been performed. The goal is to bring out the factors that are at work leading people to publish their tabs. The strategy used was to take and report what respondents said at face value, and analyze this for common themes and patterns. These are discussed next.
The most popular reason given for publishing guitar tabs online was to share the song with others. Someone who publishes tabs solely to share the song with others is simply wanting to help others. 17 people gave as their only, single motivation, to share the song with others. Of these same 17, 5 claimed they got no benefit out of publishing. These 5 are simply offering their knowledge to anyone who wants it.
Publishing tabs to avoid having to buy, or to avoid others having to buy, expensive music books is reminiscent of Bezroukov's (1999) comments about how important a common enemy is in motivating people to share knowledge. However, with only 4% of the reasons falling under these labels (one in self and one in the altruistic category) it is clearly not an important part of the tab publishing community. The guilt motivation implies that if people get free benefit from the knowledge of others they may be more likely to freely share their own knowledge.
When the self category is examined, all the reasons given, with the exception of ego and return on investment, are reasons for transcribing songs and not necessarily reasons for publishing them. 50% of respondents said they transcribed each song so they can play it themselves, then they publish. 36% claimed they do not transcribe the songs for their own purpose but do it specifically to publish and 14% claimed they transcribed for themselves and for some other reasons. Responses to question 2 (“Why do you transcribe songs?”) and question 3 (“Why do you publish the songs you have transcribed?”) are analyzed together as it was clear from the responses that many respondents answered them as if they were one question. For example many respondents, for question 3 answered with something like “see question 2.”
Once a song has been transcribed, there is in fact, not much effort involved in publishing it. All that needs to be done is that the transcription is documented in Notepad and emailed to OLGA. So for more than half of the people involved who are motivated to share the song with others, their knowledge is 1) “created” for their own enjoyment and 2) requires little effort to share. This suggests that people transcribe the song for themselves and then, as a secondary activity, publish it for others. This concept was observed by von Hippel and von Krogh (2003) in open source projects with some programmers writing algorithms for their own use and then, realizing their usefulness, submitting them to projects. However this is not seen in the bulletin board model. With the bulletin board model, an individual's contribution is of little practical use to themselves. This is an important finding — if the act of sharing knowledge increases (or clarifies, or structures, or improves or is positive in some other way) the person's own knowledge then this can be used to motivate sharing if people have a genuine interest in increasing (or clarifying, or structuring, or improving) their knowledge.
The ego reason is incorporated into theories on bulletin boards and open source projects although in each case, the factors encapsulated in ego in themselves do not fully explain the activity — no one is involved in these activities purely for egotistical reasons. However ego does seem to be an important factor as can be seen in these results and in the literature. So it can be said that if sharing knowledge results in an increase in status, people will be more willing to share. Ego appears as a motivation and as a benefit in fame in the community. This has been found in other studies of knowledge sharing (Lerner & Tirole, 2000; von Krogh, 2002).
Return on investment was not found to be an important reason for publishing, although it is closely connected with guilt. The difference between the two is return on investment is publishing so that in the future the tabs of others can be used, and guilt is publishing because in the past, the tabs of others have been used.
The pleasure motivation suggests people will be more likely to share their knowledge if they have a love of their subject. This is similar to Rheingold's (1994) suggestion that people contribute because they have an interest in the subject domain and to McLure, Wasko, and Faraj's (2000) finding that people are willing to contribute because they are “interested in maintaining the community or profession as a whole.”von Hippel and von Krogh (2003) also suggest that open source coders contribute in part because they enjoy writing code.
None of the benefits that were reported by more than one respondent would have fallen into McLure et al.'s (2000) tangible returns category. Satisfaction and fame come under their intangible returns category and feedback would fall into the community interest category. 23 of the 72 respondents claimed they received no benefit from publishing tabs. Referring to benefits, one respondent claimed he got “none. others benefit. i benefit from their work.” (Benefiting from the work of others is not a benefit from publishing tabs as anyone who has not published can still use others' work.) The most common motivation given by respondents who listed none as a benefit was to share the song with others. The next most common motivation was to improve guitar playing. (Of all the motivations listed by these respondents, 29% were to share the song with others, 24% were to improve guitar playing. The next most common answer made up just 9% of the list.)
In addition to finding support for a private-collective model of tab publishing, these results lend support to some of McLure et al.'s (2000) conclusions that people contribute knowledge to online communities because it is fun (they get fun out of the song and want to share this fun with others) and because they have a sense of moral duty (this is seen in the altruistic category reason guilt). A third reason not found by McLure et al. (2000) was status in the community although as already stated, this has been observed by several other researchers. Some support was found for the scientific review model of why people contribute. Getting feedback on the tabs was never given as a reason for publishing tabs, although 22% claimed that was a benefit from publishing. Status [which Lerner and Tirole (2000) claim is essential in open source projects] was an important motivator as seen in ego, and benefit as seen in fame in the community. Hardly any evidence was found that tab publishers are frustrated with an imperfect existing solution [which was also quoted by Lerner and Tirole (2000) as important in open source projects], although an insignificant number of respondents were annoyed by the expense of professionally published guitar sheet music. As mentioned earlier, these professional publishers would also equate to the common enemy Bezroukov (1999) talked about. Support was found for Raymond's (1998) suggestion that open source programmers contribute simply to give a gift to the community. This was by far the most common motivator for tab publishing. As for Rheingold's (1994) two suggestions for why people contribute, support is found for the second (they have an interest in the domain of the community) but not so much for the first (to receive emotional support or intellectual companionship) with 62% of respondents not (knowingly) interacting with other members of the community.
Conclusions
These results offer the following implications for organizations and online communities wishing to encourage knowledge sharing:
Enjoyment seems to be an important factor in motivating knowledge sharing. Perhaps people who get enjoyment out of the domain will be more likely to share their knowledge of it. This is made explicitly clear in the pleasure classification, but also can be seen in the desire tab publishers have to share music with other people. Additional research is needed to test this relationship. In any case, there is little an organization could do to exploit this — either employees enjoy the domain that they work in or they do not.
People who feel part of a community and get satisfaction from being part of a community will be more likely to contribute. This feeling can be cultivated by ensuring an atmosphere of loyalty and trust, and the feeling among members that they actually belong to a community and that it is worthwhile.
The act of sharing knowledge should be as close to effortless as possible to encourage contributions. For businesses this means that there should be a mechanism for sharing knowledge; popularly this would be the company's intranet, and it must be very easy to publish material on it. “Easy to publish” is difficult to quantify but effectively this probably means that the uploading of files should be hassle free and that knowledge sharers should not have to create HTML code beyond perhaps selecting a menu option such as “save for intranet.” It also means that classifying the knowledge should not be a burden. In other words, deciding where on the intranet the knowledge is to go should not be difficult. In fact, a good search facility on the intranet may make this easier (Google can be used to search intranets). It might be best if a knowledge sharer publishes their knowledge on their personal page. This means they do not have to think at all about where on the intranet their knowledge should go. Having a good search engine will mean another employee will be able to find it wherever it is. Alternatively a firm might employ an intranet Web master to upload and organize the files. All of this will help keep the publishing as effortless as possible.
The act of preparing (collecting, collating etc.) the material to be shared should have meaning in itself for the person who is preparing it. This was seen in the fact that most self motivations were motivations to transcribe a song and store that transcription in electronic form, and not motivations to publish the tabs. The publishing came later and all that publishing involved was emailing a file to OLGA. This result could be used by organizations with a little imagination. For instance, the organization might ensure that it is possible to publish knowledge in a variety of formats. An example, say annotated diagrams, might clarify this. Perhaps an architect has made some hand written notes to herself on a floor plan. The notes might be a record of changes that are necessary and why they are necessary. If the architect thinks that this is worth sharing, it should be possible for her to easily publish a copy of the plan including the hand written comments. Perhaps the floor plan could be scanned and published in jpeg format. The point is, in this example, the architect should not have to write a whole new document explaining this knowledge unless he or she really wants to.
If the act of sharing leads to increased status in the community people will be more likely to contribute. This came out in the motivations and benefits, and in the literature review.
Finally, to encourage knowledge sharing, those who make use of the shared knowledge should be encouraged to give positive feedback to the person who shared it. This could be done very easily. At the least, each month a report could be generated from the intranet's log summarising the number of times each document has been viewed. At the other end of the scale, users could be encouraged to email publishers to give them information on how their published knowledge has been helpful.
This paper has examined open-ended survey data about the motivations for and benefits from sharing guitar tabs online. To date, there has been little empirical work examining online posting forums. Some implications of the results for encouraging online knowledge sharing have been discussed.
Footnotes
All quotations are taken directly from the written responses —“sic” has not been used where there have been mistakes made by respondents.
This is an estimate. Some responses gave “college” as the highest qualification which might or might not be a degree, and the term degree indicates a different level in different countries.
Appendix
Appendices
Appendix 1.
How do you choose which songs to transcribe?
Do you transcribe songs specifically to publish them online? If you answered No, why do you transcribe them?
Why do you publish the songs you have transcribed?
What motivates you to spend time and effort doing this?
What benefits do you get out of publishing the songs?
Do you interact with the people who make use of your tabs? If you answered Yes, in what ways? Could you include some example emails with the names removed? Are the messages ever negative or nasty?
Do you keep a record of which of your tabs people are looking at? If Yes, does this affect which songs you transcribe?
How does it make you feel knowing that people are looking at your tabs?
Do you interact with other people who publish guitar tabs? If Yes, in what ways? Do you get criticism of your tabs? Do you give criticism of their tabs? How does this affect you? Could you include some example emails with the names removed?
Do you use other people's tabs? If Yes, what for, exactly?
How long on average does it take you to transcribe a song?
Personal details
Age:
Highest academic qualification:
Sex:
Occupation:
Country where you live:
Appendix 2
The following is one of the responses to the questionnaire.
Dear [AUTHOR'S NAME REMOVED], here are my answers to your questionnaire:
How do you choose which songs to transcribe? I usually pick the songs that I'd like to play along with my band.
Do you transcribe songs specifically to publish them online? No, never. If you answered No, why do you transcribe them? I transcribe them for personal use, I usually forgot a song if I spend more than a few months without playing it.
Why do you publish the songs you have transcribed? I send them to web pages because maybe they could be of some use for someone.
What motivates you to spend time and effort doing this? Well, when I began playing guitar I learnt a lot of songs from the work that someone else did by publishing his/her own transcriptions on the Internet. It's a way of returning the favour.
What benefits do you get out of publishing the songs? Absolutely none, in fact I'm receiving lots of spam (and some viruses), since this email account is now public.
Do you interact with the people who make use of your tabs?
Some people has emailed me a couple of times asking questions about the transcriptions, that is. If you answered Yes, in what ways? Could you include some example emails with the names removed? Are the messages ever negative or nasty? I'm afraid i deleted those messages but they used to be like “which kind of sound effect do you use to play that song?”, “are you sure about the chorus?” or even “the tab you made is perfect, man!” No, the messages I received never were nasty.
Do you keep a record of which of your tabs people are looking at? No, I don't track that sort of things. If Yes, does this affect which songs you transcribe? If I did that, it wouldn't affect the songs I choose to transcribe. It depends just on my personal taste. How does it make you feel knowing that people are looking at your tabs? It's okay for me, specially when I see that some of my transcriptions are circulating through the Internet and now can be found in several websites, although I originally posted them just to one site.
Do you interact with other people who publish guitar tabs? Yes, just by email. If Yes, in what ways? Do you get criticism of your tabs? Do you give criticism of their tabs? How does this affect you? Could you include some example emails with the names removed? Sometimes I've sent emails suggesting the revision of the transcription, since it was plainly wrong. But most of the times my emails just gave thanks to the author of the tablature.
Do you use other people's tabs? Yes, sometimes. If Yes, what for, exactly? I like to play a lot of songs from different bands and I begin by looking for a transcription of the song.
How long on average does it take you to transcribe a song? It depends on the difficulty of the song, but no more than thirty minutes. But if the song includes long or difficult solos it can take me even three or four hours.
Personal details
Age: 25
Highest academic qualification: Degree in Politics and International Relations.
Sex: Male
Occupation: Student/unemployed.
Country where you live: Spain Greetings and good luck with your research.
[RESPONDENT'S NAME REMOVED]
This respondent was coded as giving three motivations:
In the self category, documentation: “I usually forgot a song if I spend more than a few months without playing it.”
In the altruistic category, guilt: “when I began playing guitar I learnt a lot of songs from the work that someone else did by publishing his/her own transcriptions on the Internet. It's a way of returning the favour.” The respondent is saying here that they feel they should contribute because they have used others' contributions.
In the altruistic category share: “maybe they could be of use for someone.” The respondent was coded as giving one benefit which was none. In fact, the respondent claims they get drawbacks from publishing, although this does not stop them continuing to publish.: “Absolutely none, in fact I'm receiving lots of spam (and some viruses), since this email account is now public.”
