Artefacts of the Collective Unconscious is an ongoing digital collection of data about dreams, to which anyone is free to browse or contribute. The collection is inspired by the concept of social dreaming: the idea that dreams can tell us not just about the individual dreamer, but about our collective experience of the world: that by sharing dreams we create pictures and patterns that can illuminate concerns and hopes about about our future together.

As well as collecting and collating dream data, we represent the data in visual form using what we have called an 'oneirogram'. This is a force diagram which uses the AlchemyAPI's natural language processor to analyse themes and concepts in the dream text.

In the future, we aim to allow integration of dream data with other applications and platforms which are used to record and store dreams.

Curators: David Patman and Noah Pedrini
Want to know more?

Why are you doing this?

We're interested in the social aspect of dreams - what they can tell us about experiences we have in common, rather than just our personal history—inspired by the concept of social dreaming. We are also interested in exploring interactions between computers and non-conscious states of being, such as dreams: e.g. how dreams might (or might not) be computational and also how computing and interacting with computers might be made dream-like (Baumer and Tomlinson, 2008).

Isn't my dream just about me?

Part of your dream is likely to be about your personal experience that is unique to you, and only you will be able to understand that part. However, part of your dream will also be about the experiences that you share with other people: e.g. your family, your friends, colleagues. We hope that by looking at lots of dreams that we'll be able to see common themes and patterns emerge.

What is an oneirogram?

We invented what we call an oneirogram as a way to visualise connections between the text record of dreams and a computer-generated 'collective dream'.

The collective dream appears as the central node and is created in real-time from the text of dreams displayed for the user-selected date range. For each dream, the number of sentences extracted and the decision of which sentences to take is based on a uniformly-applied set of rules.

The connections between dreams are made using the free AlchemyAPI natural language processor. Alchemy analyses a text string and identifies concept tags using algorithms and a lexical database. Connections, or 'edges', are drawn from a dream 'node' (solid colour) to a node representing the Alchemy tag (dotted outline). The size of each node is proportional to the number of connections. Thus, dreams and tags which are most 'influential' are shown as more prominent in the oneirogram.

The oneirogram can currently show dreams for specific date ranges, and we will add functionality so that users can select locations, tags and other data.

How are connections made?

We aggregate the text of all dreams for the selected parameters (date, location, tags) and determine thematic tags for the 'combined dream'. The combined dream tags are then compared with tags for the individual dreams. If a dream shares a tag with another dream, a line (or 'edge') is drawn between them. The more connections a dream or tag has, the larger its node is shown.

How will you use my dreams and my personal information?

We will store your dream data in our digital collection, but no personal information. However, anyone will be able to see the text of the dream you record, so you may wish to disguise any details that you think could identify you or others.

Can I use your site to keep a personal dream journal?

We plan to add functionality so that users can tag individual dreams. So, if you tag each of your dreams with a unique identifier, you will be able to search for and visualise just your dreams.

Can I edit my dreams once they have been recorded?

At this point, no. We would need to create a personal account to enable you to do this. We may consider doing this in a future release. Please contact us at if you need a dream removed, e.g. if you have inadvertently recorded identifying information about yourself or someone else.

What technology do you use?

When a user submits a dream and other information on Artefacts, we store it in a MySQL database along with other data we automatically collect, such as geo-location based on IP address. We also analyse the dream text using the free AlchemyAPI natural language processor and store the returned concept tags.

When a user chooses paramaters for the visualisation (e.g. date range), we use an algorithm to generate a 'collective dream' in real time, which is also analysed using Alchemy. Using the stored tags, we create a JSON file representing node-link relationships between all of the dreams. This data file is then used to drive the visualization.

The visualization itself is a force-directed layout built using the D3js javascript library, and the image export feature leverages the canvg library. A few of the other javascript libraries we use include jQuery, jQuery UI, and tipsy.

Whose behind it?

David Patman is an engineer, social theorist and management consultant specialising in the study of unconscious group and organizational dynamics. David uses creative and associative techniques to help client groups become aware of and manage the underlying pressures that affect how they go about their work. David has a keen interest in how the spread of electronic communication technologies is changing what organizations need to do in order to survive and thrive. David divides his time between Melbourne, Australia and Hobart, where he grew up within cooee of MONA.

Noah Pedrini is a digital artist interested in exploring the changing face of community and the self in contemporary society. Anthroposts, a net art installation featuring found notes and the recordings of crowd-sourced workers reading them, was a finalist in the Venice-based Arte Laguna awards, and has been recognized by Communication Arts and other leaders in design and data visualization. His projects have been used in course materials for schools including the Maryland Institute of Art and UNC Chapel Hill He currently resides in Brooklyn, NY.

We would like to thank the development communities who contributed to the open-source projects used in this project. We'd also like to thank and Henry Ryder and the rest of the designers at The Noun Project for letting us use their icons.

Did you know that 'artefact' is spelled wrong?


Brad said:
Very interesting piece of work. I really think this could be extended to provide some useful social applications and will be in touch to talk further. In the meant time I might experiment by uploading some dreams to see what happens.

It would be helpful however to understand a bit more about the algorithms selecting the linkages.


Kelcy Davenport said:
Hello, I just wanted to say how much I enjoy your site and idea. I intend to upload dreams daily for the next 3 years as a small part of my PhD in Fine Art. Is this ok with you? And, will the website be here indefinitely?

Thanks very much for your help.

Kind regards