A pattern language for knowledge co-creation

By Yuko Onishi

Yuko Onishi (biography)

How can pattern language be used to share tips for knowledge co-creation in transdisciplinary research? What is pattern language?

Pattern language

Pattern language is an idea that originated in the field of architecture and city planning in the 1970s. The American architect Christopher Alexander and his colleagues created a common language, referred to as pattern language, that can be used by non-experts to participate in the process of city planning and building design.  

In this pattern language, the rules of thumb for solving common and timeless problems in design are summarised in units called ‘patterns.’ Each pattern describes a specific problem, the situation or context in which it likely occurs, and the core of the solution to that problem.

The solutions are not written as specific procedures or manuals, but rather as ‘hints’ for solving the problem. Therefore, the solution can be used in many ways based on one’s own needs and situation. 

Each pattern is assigned a name, so it can be used to refer to a particular combination of ‘problem statement,’ ‘context description,’ and ‘proposed solution’ when talking with others. The fact that patterns can be used as ‘words’ and can be linked to each other is the reason why a set of patterns is called a ‘language.’    

Pattern languages have subsequently been created in other fields such as education and healthcare. For their use in modelling, see the i2Insights contributions by Scott Peckham on Looking for patterns: An approach for tackling tough problems and Sondoss Elsawah and Joseph Guillaume on Sharing integrated modelling practices, both Part 1: Why use “patterns”? and Part 2: How to use “patterns”?

Pattern language for knowledge co-creation

Colleagues and I are developing a collection of patterns to guide knowledge co-creation in transdisciplinary research and to overcome common challenges, as described in the four patterns to follow and as replicated in the image below.

Expressing lessons learned from dealing with such re-occurring problems in the field as patterns can create a pattern language for transdisciplinary research.

Pattern 1: Communicative language.

The problem: Scientists often unconsciously use jargon or terms with a specific meaning. Communication troubles may arise because non-scientists cannot understand the terms in the same way as the scientists.
The solution: Use plain language when communicating with stakeholders and avoid scientific jargon. When this is not possible, explain the meaning of the jargon or term and why it is necessary to use it.

Pattern 2: Mind the difference.

The problem: Even within the same country, differences in culture, lifestyle, thought styles, and values exist.
The solution: Researchers entering a field site, for example a community, should recognise these differences and adapt to the situation.

Pattern 3: Make use of skills.

The problem: Without trust, stakeholders are unlikely to share honest opinions or provide key information.
The solution: Do what you can do for the community, rather than focusing on your research topic. Sincere attitudes and actions will help build trust with the community.

Pattern 4: Enjoy exploring new ideas together.

The problem: Researchers think that they are unable to find solutions to the problem and frustrated that progress is slow.
The solution: It takes time to find innovative ideas and solutions. Be aware that this is a time-consuming process and try to enjoy the exploration.

Four patterns for knowledge co-creation in transdisciplinary research (Source: Yuko Onishi).


Are any of these patterns useful or applicable in your research context or regions? Does this i2Insights contribution highlight any ideas about patterns in transdisciplinary research for you?

To find out more:

Co-creation Project. (2021). Tips to foster knowledge co-creation and guide transdisciplinary research (TDR). (Online – webpage): https://cocreationproject.jp/en/learn/tips/

Biography: Yuko Onishi PhD is an assistant researcher at the Research Institute for Humanity and Nature, Kyoto, Japan. Her interests are in transdisciplinarity, co-creation concepts and societal outcomes.

19 thoughts on “A pattern language for knowledge co-creation”

  1. Thanks Yuko for your explanation. In one experience i had, I found it useful to mine the seeds from literature and focus the discussion of the group on contextualising the pattern (e.g. constraints,…etc). This may help move the discussion faster and set the focus on fleshing out the details.
    Thanks again. I am excited to follow your work.

    • Dear Sondoss, thanks again for your reply. I also think that it would be faster in some cases to have some seeds from literature. However, there also is a danger that we miss out some of the very unique ideas when the discussion starts from a set of seeds (that are presumably good ones considering that they are already published).

      Since I had already spent lots of time, interviewing people from scratch, I would like to believe that it was worthwhile. However, it was a very long and challenging process, so when I have the time for, I would like to review the process and find out a better way in the future.

  2. Dear Yuko,
    Thanks for sharing your interesting work. I believe that patterns and pattern languages represent great artifacts for knowledge sharing and for helping to tackle complex issues that require interdisciplinary efforts. I think the users of the patterns you are suggesting can benefit a lot if you included the constraints that might limit the application of the proposed solutions as part of the pattern. Also, are these patterns interdependent? i.e. does applying any of them affect applying the others? or does the order of applying them affect the overall knowledge co-creation process?

    • Dear Ebrahim,

      I agree completely that it is important to include the constraints. Thank you very much for a nice suggestion.

      Some patterns are closely related but in general (or as far as I can remember), they are not interdependent. Some patterns are useful at an early stage of research projects, and others are related more at the later stage (e.g. producing outputs). I am currently trying to classify the patterns, so I can provide more concreate answers to your question when I finish. (Some patterns are useful at any stage as well, so it is not straightforwad.)

      • I love how beautifully the example patterns are presented!

        Regarding constraints, I just wanted to agree with Ebrahim and Sondoss about the importance of contextualisation.

        The definition I tend to use is that “A pattern describes how a practice is a possible solution to a commonly encountered problem in a context”, and the words “possible” and “in a context” are probably the most important to me.

        Without them, there is a risk that the solution is considered universal, whereas a strength of the pattern concept compared to e.g. a protocol or recipe is the ability of the user to make the pattern their own and think about whether and how it applies to them.

        • Thank you for raising an important point! In the past few months, I have revised our patterns for co-creation and the latest version now consists of a set of “context”, “problem” and “solution”.

          I excluded the “context” in an earlier version (i.e. the figures in this webpage) because it was difficult to identify an appropriate context for all of the patterns (and make sure that each pattern is linked to unique context).

          As this has been very tough and time-consuming work, I was very happy to see your comment, pointing out the importance of context. So, thank you so much!

          I would like to share the new patterns with you once I managed to put them in a good shape.

  3. Dear Yuko, many thanks for sharing your work. I am inspired by The Pattern Language as well. Similarly, I want something non-experts users can apply in diverse settings, functions, and disciplines on a daily basis. The inherited or conventional patterns tend to exclude too many people, stifle creativity, and over-control. The 33 patterns we dubbed Liberating Structures (LS; https://www.liberatingstructures.com/) specify how people are included and participate as they interact. Each LS pattern has the same minimalistic “DNA” made of five micro-structuring elements.

    Your four pattern language categories are complementary to the ten LS principles: Include and Unleash Everyone; Never Start Without Clear Purpose(s); Practice Deep Respect for People and Local Solutions; Build Trust As You Go;’ Learn by Failing Forward; Practice Self-Discovery Within a Group; Amplify Freedom AND Responsibility; Emphasize Possibilities: Believe Before You See; Invite Creative Destruction To Enable Innovation; Engage In Seriously-Playful Curiosity. Rather than informing transdisciplinary research, they guide fidelity and connection among diverse communities of users.

    I will take a closer look at your co-creation project. Appreciatively, Keith

    • Dear Keith, thank you for your comment. I read the website on LS with much interests. Yes, there is a lot of similarities. I also feel that much of the tips for TD research are common with general communication tips for collaboration. I like the icons on your website, which lead to a page with detailed explanations. I might think about doing something similar with my patterns (*^。^*)

      How are your ten principles related to 33 patterns? For instance, “trust building” is no doubt important, but I feel that it’s difficult to provide a “recipe” that works for all. I am curious to know what you have found in your experience of using your patterns to a wide variety of participants.

      • Thanks for responding. My sister in law made those playful icons. A family affair! The Liberating Structures (LS) principles arose in the midst of 10 years of action research in the field. With a complexity science lens, we were consulting with large organizations facing entangled challenges. The protocols we were prototyping were a radical shift from mechanistic “reduce things to smaller and smaller parts” approaches. Our focus on relationship patterns among the parts and distributed control revealed the assumptions and behaviors underlying conventional practice. The differences are dramatic and can be measured with developmental evaluation. See the must do’s and must not do’s on the chart on this page: https://www.liberatingstructures.com/principles/ and in this i2Insights contribution: https://i2insights.org/2021/08/31/principles-for-inclusive-groupwork/. Further, we wanted to guide fidelity and hold the community of users together as the work spread. LS are more patterns for relating than tools to centralize control. As such, they are surprisingly relevant across cultures and domains. A short answer to a deeper question. ;^)

        • Sorry for a late reply. My project actually ended officially last week, so I’ve been really busy (https://cocreationproject.jp/2023/03/31/completion-of-co-creation-project/).

          It is such a lovely story that your sister made the icons! I found that the good communication is neccessary to develop the figures for the patterns, so a family affair is perfect.

          Thank you also for explanation of the LS principles. I am currently in the process of finalising the patterns of knowledge co-creation (and now there are 30 patterns). I would like to make a closer comparison with your LS principles once I finish them.

          • Love the humility in your summary statement: “Co-creation is not always necessary or efficient means for problem solving. However, when you face a big challenge that you cannot handle by yourself, if you can work with someone else and find a way to solve it together, co-creation can be incredibly rewarding and fun.” If and when the time is right, I would be happy to continue this conversation comparing approaches. keith@liberatingstructures.com. I know two LS maestros in Japan who may want to join in.

  4. Dear Yuko, thanks for the interesting blog. Great to see how patterns are used for synthesizing and communicating lessons. I am interested in your experience of eliciting, formulating, and validating the set of patterns. Could you please shed some light on the methodology you have used, and any insights and lessons from applying the methodology in knowledge co-creation?

    • Dear Sondoss, thank you very much for your comment. To formulate the patterns, I first had a workshop with several scientists to collect the “seeds” or patterns. Then, I had to move on to online interviews due to COVID-19. After collecting more than 400 seeds, I carried out another intensive workshop (7 full-days, on site) to synthesise the seeds and develop the patterns. I then organised several workshops (online) to review the patterns in groups. My lesson learnt from this process is, albeit obvious, face-to-face workshops are much, much, better than online workshops. We invested lots of time learning and applying online tools such as Miro, but in the end, they were never more effective than on site workshops. I also found that selection of the participants is very important in the workshops.


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.