This text has appeared in Kontraste Cahier No 2
Sonic Arts Press, 2012
Dark Energy, Dark Matter, Big Data, Intimate Data
Roger F. Malina
Astronomy, with the agricultural and health sciences, is no doubt among the oldest human sciences. The regularity of events in the sky, the daily and annual cycles, provided a predictable framework in a world that was often chaotic for early hominids. Humans lived at the mercy of climactic variations and disasters, threats from predators and unpredictable diseases and events. Astronomy on the other hand provided a metronome for human existence. And unusual events in the sky took on symbolic and religions importance. For most of human history astronomy has been a ‘regalian’ science associated with the ruling classes. Its role in navigation, even to the present with the contributions of the space sciences to the GPS systems, provided a continuous link between science and economic benefit. Today China and India use the space sciences not only to demonstrate their technological prowess, but because we need the space sciences to help ‘manage’ our impact and activities on the planet.
Within astronomy, cosmology has had a particularly important role as it has contextualized our relationship to the larger universe. The Galilean revolution was at its root cultural. It is no accident that the theories of gravity have played such an important role in the development of modern science. The Universe on the largest scales provides one of our richest laboratories, testing and extending our understanding of the world around us. Our understanding of the expansion of the universe from a hot dense phase, the Big Bang, has provided an overall ‘calendar’ within which stars, planets and then life has emerged. The theory of evolution is joined at the hip with the theory of the big bang; astronomers played an important role in establishing the time scales of geological history; the cultural impact of astronomy is still felt in the debates with fundamentalist religions. Cosmology today is one of the most active areas of astronomy. New generations of telescopes, space based and ground based, are bringing a data flood of the nature and distribution of matter on all scales and distances going back to the first formation of stars in the universe after the Big Bang.
The Dark Universe
It is therefore ironic that astronomy is undergoing such a crisis of epistemology today. The oldest science finds itself among the youngest sciences, with established understandings unsettled by new data. A recent issue of Science Magazine  detailed ten fundamental areas of ignorance in modern astronomy that include:
What is Dark Energy, 95% of the content of the universe is of an unknown nature except that it is causing the expansion of the universe to expand.
What is Dark Matter: 83% of the physical matter in the universe is of an unknown nature except that it it holds together galaxies and other large scale structures in the universe.
Where is the Missing Matter: Even for the matter that we can see or deduce its presence, 50% is unaccounted.
The article goes on to detail other major areas of ignorance such as the source of the most energetic cosmic rays in the universe, or how stars explode.
In 2011 Saul Perlmutter, Adam Riess and Brian Schmidt were award a Nobel Prize in Physics for the discovery of dark energy. This Nobel Prize is perhaps ironic because it is awarded for discovering that we are ignorant of the nature of 95% of the universe. The story of this discovery is almost an exemplar of how good science is done, with small groups of scientists taking data, discovering step by step that the current understanding was very flawed and slowly convincing their colleagues who initially dismissed their work. As their work was on, it motivated the invention of new instruments and new telescopes that could obtain the data to confirm or refute their ideas. New generations of space telescopes are now on the drawing boards that would collect vast amounts of data to perhaps elucidate the nature of dark energy.
The era of big data started early in astronomy. When I started my career we were still using photographic plates. Then astronomers digitized their photographic plates. Then diode arrays and eventually Charge Coupled Devices (as used in cell phone cameras) started generating ever larger volumes of data. It was then realized that these flows of data could not be combined; each field of astronomy used different data formats, different software systems, different archiving mechanisms. The astronomical community, with the support of funding agencies, mobilized to develop data and software standards. Today with online virtual observatory data bases scientists, or citizen scientists, have access to large data sets from multiple telescopes. In a very real sense most of the telescopes are now networked into a large collective observing machine. New professions of data analysts have emerged and indeed many astronomers today have never been to a telescope or taken their own data; they use the data archives to make new discoveries. This evolution to the big data era spread rapidly to other fields of science, such as genomics, and now to business, government and the social media industries.
As many have pointed out big data is not just more data. Historian of Science Daniel Boorstin called this an ‘epistemological inversion’. What he meant was that the way that science could be done was changing. When Charles Darwin went onto his journeys to the Galápagos he was in search of new data. Data were rare; indeed all of Charles Darwin’s data that transformed our understanding of human nature fit into a series of notebooks on his study book shelf. When data becomes plentiful, it changes the way that most scientists do their work. They can study archives instead of studying the world. The way that governments fund science encourages the building of new instruments to take new data. The result is that the direction of science, which questions are investigated, and the method of science itself changes. There is often little funding to actually analyse and draw conclusions from the data. Daniel Boorstin joked at a World Space Congress that maybe space agencies should stop taking data for a while since they didn’t have time to analyse all of it.
Citizen Science and Data Rights
In the information economy, ‘data is power’. The opening of astronomical data bases to the public has led to a large growth of citizen science with private citizens able to make scientific discoveries. This citizen science movement goes beyond the concept of the ‘amateur’. In many fields citizen scientists are not only analysing online data but generating new data. The social and political ramifications are important as is being shown by the community mapping and community ‘remote sensing’ movements. With kite and balloons and remotely controlled drones, and of course with cell phones, private citizens can take data to contest what governments or companies are claiming. Last year the Buckminster Fuller prize was awarded to a number of such groups, including a Congo citizen science mapping project that was successful contesting illegal logging of old growth tropical forests. The issue here is not big data but the right data and the right to data.
Several years ago I wrote a Open Observatory Manifesto asserting two new simple rights. I first asserted that all citizens had a right to access data taken using tax payer money. In astronomy this is actually implemented in most government grants; the astronomers are required to make their data public after a certain time period. It is hotly contested in many other areas where governments refuse to make data, paid for by tax payers, available to the public. The second assertion I made was that citizens had a duty to take data and contribute it to the data commons. In fact this is rapidly happening as social media systems archive all kinds of data uploaded by private citizens. We are rapidly becoming a data taking culture. The biomedical sciences are being transformed as the cell phone becomes the generic data taking interface. Data privacy has become a major ethical and societal issue.
Data Visualisation and Sonification
When you have very little data you can look at all of it. Astronomers used to study each galaxy in an image. When you start having millions or billions of objects in your database it is humanly impossible to do this. You start by automating the process but inevitably this means you build in blind spots. Every data processing algorithm filters, sorts, selects and often throws away most of the data.
In astronomy one of the popular stories is how astronomers made surprising astronomical discoveries in data collected by the US Military. The military satellites were watching for nuclear bomb tests during the cold war era. The data analysis system rejected any signals that seemed to come from above the satellites, because they were looking for tests carried out underground or in the air. The astronomers discovered that the signals coming from above the satellite were real and were a previously unknown phenomenon now known to be gamma ray stars and galaxies.
As data volumes grow, traditional scientific illustration techniques become inadequate and this has led to the growth of new professions and techniques in data visualization. Data can be show in three dimensional immersive environments, in ways that are interactive and manipulable. Techniques in complex network science allow the structure of the data to be analysed drawing conclusions about the content of the data. Infoviz, bioviz and dataviz conferences are proliferating
More recently scientists have started sonifiying their data as well as visualizing it. Human perception functions differently in visual and aural perception, different kinds of patterns can be detected and time evolutions noted. Sonification goes beyond alarms and alerts to systems of complex data representation exploiting the 3d and time-based nature of sound. Interestingly composers and sound artists, such as Scot Gresham Lancaster, have been on the forefront of developing these techniques.
No Data or the Wrong Data
What if there is no data or we are collecting the wrong data? The dark energy astronomers are collecting vast amounts of data but since we don’t have good theoretical models, other astronomers have pointed out that the data may be useless. Without good hypotheses to test how do we know what data is relevant? The observational astronomers reply that dark energy was discovered without a guiding theory, and that many discoveries are not driven by the process of confirming or falsifying hypotheses. Science indeed advances by both approaches. During the birth of the big data transformation some argued that it was ‘the end of theory’. Just collect data, look for correlations and do extrapolations. In many cases this may work very well. But in others understanding causal networks are needed to interpret correlations and extrapolations. There is a very real danger that the data flood will blind us to the fact that we don’t have the really needed data. And its easier now to get funding to analyse big data that to fund research where you don’t yet know which data you need.
Quantitative Data and Qualitative Data
One of the ancient battlegrounds between the sciences and humanities has been whether scientific understanding can only come from quantitative data. The social sciences and the cognitive sciences find themselves straddling the digital divide. Within the arts, the digital and new media arts are still fighting fundamental battles with art forms that don’t rely on manipulating digitally encoded, quantified, data. In recent years the birth of the Digital Humanities has re-awakened these disputes as the new generation of humanities scholars, born digital, develop new research strategies that are more easily funded today than pre-digital scholarship. Big Data is re-orienting the humanities, driving curiosity towards questions that could not been tackled before, but also putting in the shade fundamental questions which have no data or rely on un-quantifiable qualitative analyses.
For artists perhaps the question is elsewhere; the human experience with an art work relies on qualia of human cognition. That experience is in a sense neutral to the technology used to develop the art work. Artists have often been early adopters of new technologies, and in many cases have made inventions for their art making that have been widely used. Metallurgy and chemistry have long straddled the fine and applied arts. We now know that the human senses are very efficient filters, and that almost all the world around us is undetectable to human senses directly. Most of the universe is dark. That artists use data obtained by scientific instruments seems to be a desirable process of cultural appropriation of phenomenon to bring them into the intimacy of personal perception and cognition. Gyorgy Kepes in his books like The New Landscape in Art and Science (1956), asserted the right of artists to use scientific data as a raw material like any other. In a sense the citizen science movement has its equivalent in ‘citizen’ art, with artists also generating new data using scientific instruments for their own purposes, I believe this should be encouraged. There is a growing movement to find new ways to cross link Science and Engineering to the Arts and Humanities. The Dark Universe and Big Data are one common ground to be explored.
1.Mysteries of Astronomy, Science Magazine, June 2012 http://www.sciencemag.org/site/special/astro2012/index.xhtml
Roger F. Malina is an astronomer and editor. He is a Distinguished Professor of Art and Technology at the University of Texas, Dallas where he is developing Art-Science R and D and Experimental publishing research. Former Director of the Observatoire Astronomique de Marseille Provence. His specialty is in space instrumentation; he was the Principal Investigator for the NASA Extreme Ultraviolet Explorer Satellite at the University of California, Berkeley. He also has been involved for 25 years with the Leonardo organization whose mission is to promote and make visible work that explores the interaction of the arts and sciences and the arts and new technologies. Since 1982 he has been the Executive Editor of the Leonardo Publications at MIT Press. More recently he has helped set up the Mediterranean Institute for Advanced Studies (IMERA) and is co chair of the ASIL ( Arts, Sciences, Instrumentation and Language) Initiative of IMERA which hosts artists in residence in scientific research laboratories of the Marseille region.