aa r X i v : . [ a s t r o - ph ] A p r Fundamentalist physics: why Dark Energy is bad for Astronomy
Simon D.M. White
Max Planck Institute for Astrophysics, Garching bei M¨unchen, Germany
Astronomers carry out observations to explore the diverse processes and ob-jects which populate our Universe. High-energy physicists carry out experimentsto approach the Fundamental Theory underlying space, time and matter. DarkEnergy is a unique link between them, reflecting deep aspects of the Funda-mental Theory, yet apparently accessible only through astronomical observation.Large sections of the two communities have therefore converged in support ofastronomical projects to constrain Dark Energy. In this essay I argue that thisconvergence can be damaging for astronomy. The two communities have dif-ferent methodologies and different scientific cultures. By uncritically adoptingthe values of an alien system, astronomers risk undermining the foundations oftheir own current success and endangering the future vitality of their field. DarkEnergy is undeniably an interesting problem to attack through astronomical ob-servation, but it is one of many and not necessarily the one where significantprogress is most likely to follow a major investment of resources.
The pursuit of a deeper truth, of a fundamental theory which underlies all others, isa powerful motivator in physics. So too are curiosity and awe at the richness of Nature,at the connectedness which allows disparate and seemingly unrelated processes to produceorder, beauty and diversity from apparent chaos. The first motivation is, perhaps, mostevident in high-energy physics, where a “theory of everything” has periodically appearedwithin reach, occupying many of the most talented theoreticians. The second is evident inmore interdisciplinary, less “fundamental” fields, solid-state physics, evolutionary biology,or astrophysics. Fundamentalists prize the depth of their research, seeing it as a means toabstract from the complexity of the world a Truth which embodies the ultimate foundation ofthe physics of particles and fields, thus, by extension, of all physics, chemistry and biology.Generalists, on the other hand, prize breadth and interdisciplinarity which promote theperception and appreciation of the many truths underlying complex phenomena. In theirview, the fundamental theory of everything will contribute nothing to our understanding ofthe origin and nature of life.The discovery of Dark Energy, a near-uniform field which appears to dominate theenergy density of the current Universe and to drive its accelerated expansion, has led astro-physicists and high-energy physicists to make common cause. The apparent properties of 2 –the Dark Energy, in particular the extremely low associated energy scale, are entirely unex-pected in the standard model of particle physics and extensions such as supersymmetry. Thissuggests to many that Dark Energy may somehow reflect the unification of gravity with theother fundamental forces, and hence, paradoxically, physics at energies far above those thatcan be probed directly with accelerators. At present it seems that the properties of Dark En-ergy can be explored only through astronomical observations, in particular through precisemeasurements of the recent expansion history of the Universe and of the growth of cosmicstructure. Such measurements require observation of large samples of complex astronomicalobjects such as galaxies, galaxy clusters and supernovae. In consequence, astronomers inter-ested in supernovae and in cosmic structure formation have been working intensively withhigh-energy theorists and astrophysical cosmologists to design projects which might achievethe required precision. Such collaborations bring risks as well as benefits because of differentmotivational backgrounds and different methodologies in the two communities. It is on thisissue that I wish to focus in this essay.
HST and WMAP
A useful illustration of the contrasts between the motivation and the modus operandi ofthe two communities is provided by two current satellite telescopes: the Hubble Space Tele-scope (HST) and the Wilkinson Microwave Anisotropy Probe (WMAP). These two projectshave very different scale and duration (in particular, HST is well over 10 times as expensiveas WMAP) but this is not the aspect of the missions which concerns me here. Both havebeen extremely successful and have had very substantial impact both in the professional sci-entific community and among the scientifically minded community at large. They can serveas exemplars of supremely successful projects driven primarily by traditional astrophysicistsin the case of HST, and by fundamental physicists and cosmologists in the case of WMAP.HST, a NASA-led collaboration between the North American and European astronom-ical communities, was planned over many years as the first true observatory in space. It wasanticipated that it would make great advances because its location above the atmosphereallows observations at higher resolution, at shorter wavelengths, and with less contaminat-ing sky emission than is possible with ground-based optical telescopes. Studies made beforelaunch assessed how these capabilities might advance research frontiers as perceived at thattime, but a clear expectation was always that opening new windows of observational param-eter space would trigger new and unexpected discoveries in areas where little was previously 3 –known. By and large this expectation has been borne out. HST is now primarily known fordiscoveries which were not part of its original science case.WMAP was proposed by a small and tight-knit group of scientists in response to aNASA call for suggestions for a mid-sized mission. All the proposers had a strong track-record of instrumental development for studies of the Cosmic Microwave Background (CMB),and several of them had worked on NASA’s previous CMB satellite COBE. Over the period1989-1993, COBE had demonstrated the near-perfect black-body nature of the CMB andhad detected the weak temperature variations predicted 20 years earlier to remain as visibleechoes of the early fluctuations from which all present structure grew. (The PI’s of the tworelevant instruments were awarded the 2006 Nobel Prize in Physics for these discoveries.)Theoretical predictions for structure in the CMB are robust within the standard paradigm.Detailed measurements should not only test this paradigm, but should also give precisemeasurements of the geometry of the Universe, of its contents in ordinary matter, darkmatter and dark energy, and of some aspects of the process which created all structure,perhaps during a very early epoch of cosmic inflation. The WMAP team designed theirexperiment specifically and optimally (given available technology) to map fluctuations ofthe kind predicted by theory and measured on large angular scales by COBE. After severalyears of operation they have now successfully achieved this goal.One way to contrast the nature of these two projects is through the images which havebecome emblematic of their success. In figure 1 I show two of the best known images fromHST, the picture of a region of the Eagle Nebula often referred to as the “Pillars of Creation”and the very deep exposure of an apparently blank piece of sky known as the Hubble DeepField. Both of these pictures have had enormous public and professional impact, achievingiconic status. Both would be considered beautiful by many viewers. To my mind theirbeauty lies in the complexity of the structures and in the way they resonate with visuallyfamiliar images but in a new and striking context. The Pillars of Creation are reminiscentof backlit thunderclouds, evoking the reappearance of sunlight after a storm, yet they depictthe hidden birthplaces of stars. The Hubble Deep Field shows us galaxy images almost likethose in coffee-table photographic atlases, yet these apparently neighboring systems in factstretch back through time nine tenths of the way to the Big Bang. In both cases we see richand complex systems whose structure and evolution are evoked rather than characterisedby the images. Quantitative analysis is possible, but it serves to construct approximatephenomenological models rather than to measure well-defined physical parametersThe best known images from WMAP are shown in figure 2. One is a rendition of theCMB sky with all foregrounds removed and with a dramatic colour code to emphasise the 4 –tiny temperature fluctuations detected by the satellite. This has no resonance with the famil-iar world, appearing more like a mathematical construction. To a very good approximationit is a random noise field and the statistical properties of its several hundred thousand pixelsare adequately described by a six-parameter model. The second image compares the angu-lar power spectrum of this sky map with a prediction based on an a priori physical modelwhere all six parameters have been adjusted to values which are fully compatible with thoseexpected from independent, non-CMB data. The agreement is a triumphant affirmationof the power of physics as a description of our world. Although the present universe maybe complex, the early universe was simple, and we can calculate the statistical propertiesof its structure from first principles. Fitting the observed data puts tight constraints on“fundamental” properties of the universe such as its overall geometry, its contents in darkenergy, dark matter and ordinary baryons, and the process from which all structure origi-nated. These properties affect the later growth of galaxies and stars, but the CMB sky offersno insight into the complex regularities which characterise these systems.The following table gives another view of the contrasting properties of HST and WMAPwhich illustrates some of the differences in scientific culture which concern me in this essay.HST WMAPAn observatory An experimentDesigned for general tasks Designed for a specific taskServing a diverse community Serving a single, coherent communityProgramme built through proposals Programme set at designMany teams of all sizes A single moderately large teamMany results unanticipated Main results ‘planned’Nourishes astrophysics skills Nourishes data-processing/statistics skillsPublic support as a facility Public impact through resultsMost of these contrasts are self-explanatory, but the last one may deserve more comment.In the wake of the Columbia disaster the NASA administration decided that the plannedshuttle mission to service the HST was too risky and the telescope must therefore be allowedsuccumb to its natural degradation in orbit and instrumentation. This caused a tremendousoutpouring of support, not only from almost the entire astronomical community, but alsofrom the media, from the general public, and from the astronauts themselves. Largely inresponse to this, the servicing mission is again on the NASA roster. Although the impact ofWMAP’s results was enormous, it seems unlikely there would have been such an emotionalground-swell of support had NASA decided to discontinue its operations after four years. 5 –This broad public affinity for astronomy reflects widespread interest in deep questionsof origin and fate which earlier civilizations addressed through creation myths. Similaremotional undercurrents explain the preponderance of ‘space’ themes in popular sciencefiction and the remarkable world-wide community of amateur astronomers. The latter unitesenthusiasts across generations, across skill levels, across social strata, and across nationaland cultural boundaries. Amateur astronomers build their own telescopes, use them todo research of significant if not forefront interest, and maintain a lively and high-qualitymagazine literature featuring substantive reviews of new results from professional research.Astronomy resonates with the popular imagination through its combination of complexityand regularity, of the familiar and the strange, as well as through its extraordinary andseemingly limitless range of subjects for study, from the beginning of time to the birth ofstars, from the peculiarities of black holes to those of planets, from the origin of the elementsto that of spiral galaxies, from dark energy to the preconditions for life. The fact that itis hard to imagine an enthusiastic amateur community devoted to high-energy physics isanother indicator of the cultural differences between the two fields.
The two cultures
Astrophysics and high-energy physics have a number of common features. Neither hasany direct application to everyday life, even if their instrumental and computational needssometimes lead to significant technological spin-offs. Both deal with phenomena on scaleswhich differ vastly from those of normal human experience. Both require very expensiveequipment. Despite this, the research communities in the two fields differ notably in theirattitudes, in their motivations, in their modi operandi , and in the value systems by whichthey judge their work.Astrophysics aims to understand the structure and behaviour of inherently complexsystems and as a result is interdisciplinary and synthetic in character. An intuitive feelingfor the interplay between phenomena from many areas of physics is needed, for example, tomodel the formation of a galaxy. High-energy physics, in contrast, is reductionist, aiming tobreak phenomena down into ever more fundamental and more abstract entities, discardingalong the way complexities which may mask the underlying Truth. Thus astrophysiciststend to be generalists, prizing breadth of knowledge, while high-energy physicists tend tobe specialists, prizing the depth to which they probe the underlying structure of matter.In experimental work astrophysicists seek many truths associated with many phenomena,and the best forefront research is characterised by diversity and opportunism. In particle 6 –physics the quest for the fundamental Truth has led to a focus on a much smaller number of‘important’ questions (the origin of mass, the unification of quantum mechanics and generalrelativity...) and to the organisation of industrial-strength teams to address them. Newinsights in astrophysical research appeal on many levels, intellectual, emotional and aesthetic,but they rarely display the quasi-mathematical rigour of major advances in particle physicssuch as the understanding of asymptotic freedom or of the Higgs mechanism. Astrophysicistsare universalists, democratic in perceiving interest in all aspects of the cosmos, while high-energy physicists are fundamentalists, cleaving to the pursuit of the single Truth.Many of these differences can be traced to the fact that theory has traditionally beentested against reality through controlled experimentation in high-energy physics, but throughobservation in astrophysics. The remoteness and scale of astronomical systems precludecontrol of initial or boundary conditions, while long timescales make evolution unobservablein most individual objects. Astronomers are forced to work with “snap-shots” of non-ideal,strongly interacting and complex systems. This has produced a research strategy quiteunlike that in fields where experimentation is possible. When planning major new astronomyfacilities, the principal design drivers are usually:1. to complement and extend previous facilities;2. to maximise the discovery potential; and3. to minimise the risk of scientific failure.The emphasis is on enlarging capabilities by opening previously unexplored regionsof observational parameter space (in wavelength, angular resolution, sensitivity...) ratherthan on targetting a specific scientific issue. The science case, for HST, as for most majorobservatories, was based on a wide range of problems from many areas of astrophysics. Theastronomical community has, nevertheless, always considered HST’s principal value to be theavailability of most of its observing time for programmes proposed after launch by individualresearch groups. Most astronomers no longer remember the original science justification forHST or most of the Key Programmes implemented to address it.To some extent, these considerations also apply to the design of major facilities for high-energy physics, but even a global facility such as the Large Hadron Collider is only able toaddress a relatively narrow range of problems and to conduct a small number of experiments,each carried out by a large, international team of physicists. These experiments are set uplargely according to traditional physics methodology: 7 –1. identify the potential capabilities of new instrumentation;2. identify issues that these capabilities might address;3. refine the designa) to address the important issues optimally,b) to exclude confusing factors.Team members specialise in optimising particular aspects of the experiment (magnets, detec-tors, data analysis...) and may work for decades before seeing data. Such long-term effortsrequire structured and hierarchical management, and few physicists outside the teams areable to work directly with the data. This contrasts with HST where science is primarily car-ried out through programmes that last a couple of years from proposal to completion and areindependent both of the instrument teams and of the science case which justified instrumentconstruction. The HST model offers young scientists a much wider range of opportunitiesfor scientific creativity and visibility than most major accelerator experiments.
Dark Matter and Dark Energy
Over the last two decades a standard paradigm has emerged for the evolution of cosmicstructure. One of its most striking aspects is the assertion that the current universe isdominated by two unexpected and apparently independent components, Dark Matter andDark Energy. The need for unseen matter to explain the dynamics of galaxy clusters wasfirst pointed out by Fritz Zwicky in the 1930’s, but only the last 25 years have seen wideacceptance of the idea that cosmic structure growth is driven by a gravitationally dominantpopulation of some new kind of weakly interacting particle. General acceptance of the ideathat the current expansion of the universe is accelerated by some form of Dark Energy iseven more recent, although the Cosmological Constant was introduced by Einstein as partof his theory of gravitation and is a viable explanation of current observations.Both Dark Matter and Dark Energy are seen as fundamental by high-energy physicistsas well as by astrophysicists. All currently viable elementary particle candidates for darkmatter require an extension of the standard model of particle physics with the lightestsupersymmetric particle being, perhaps, the current favorite. If this were confirmed, itwould prove that the early universe was sufficiently hot that supersymmetry was unbroken.Dark energy seems to require an even more radical extension of current theories, perhaps aunification of quantum mechanics and general relativity in some form of superstring theory. 8 –The current evidence for both dark components is purely astronomical, and it appears that only astronomical observations provide a means to constrain properties of Dark Energy.Thus, the experimental testing of the hottest idea in current high energy physics dependsto an unprecedented degree on astronomers, and the two communities have collaboratedsubstantially in planning major new initiatives to address the issue.From an astronomical point of view, however, the Dark Matter and Dark Energy prob-lems differ qualitatively in their richness and in their interaction with the rest of the field.Dark Matter drives the formation of galaxies and galaxy clusters and influences all aspects oftheir structure. Its distribution can be mapped directly using gravitational lensing, and canbe inferred indirectly both from the dynamics of galaxies and intergalactic gas, and from thestructure of fluctuations in the microwave background radiation. The current favorite candi-date, the lightest supersymmetric partner of the known particles, should produce annihilationradiation which could be imaged by planned gamma-ray telescopes. Dark Matter may soonbe observed directly by underground “telescopes” which are rapidly improving their abilityto measure the occasional collisions of Dark Matter particles with ordinary matter, and itmay be detectable in experiments at the Large Hadron Collider. Dark Matter studies thusimpact directly on most aspects of extragalactic astronomy and astrophysical cosmology, aswell as stimulating astroparticle experiments and research programmes at accelerators.In contrast, Dark Energy studies have little or no impact on other areas of astrophysicsand experimental high-energy physics. Models have been proposed in which Dark Energyinteracts with Dark Matter, resulting in observable effects on structure formation, but inmost models the two components are effectively independent of each other. The effects ofDark Energy are then manifest only in the overall expansion history of the universe and inthe linear growth rate of irregularities. If Einstein’s theory of gravity holds, one of thesefunctions can be derived from the other and all astronomically accessible information aboutDark Energy is then contained in a single observable function, the expansion rate as afunction of cosmic time. Current data are all consistent with the expansion history expectedif Dark Energy behaves like a Cosmological Constant. Estimates of the current value ofthe relevant dimensionless parameter are in the range w ∼ − ± .
1, where w = − So why is Dark Energy bad for astronomy?
I come now to the crux of my argument: how the current emphasis on Dark Energy asa principal driver of astrophysical research can undermine not only the methodological basisof astrophysics, but also its attractiveness to its best practitioners, to the most talented ofnext-generation scientists, and to the public at large. In my view, such negative consequencescan result from importing the alien culture of high-energy physics, especially in combinationwith an independent trend towards “Big Science” which is currently afflicting astronomy.The dangers I see are of three kinds: inappropriate risk assessment in the design ofmajor programmes; investment of scarce resources in programmes which do not enable newastrophysics or promote advances over a broad front; promotion of a fundamentalist valuesystem and a managed work culture which will make astronomy unattractive to the brightest,most creative and most ambitious young scientists. Let me discuss these in turn.The remarkable advances made recently through studies of the microwave background 10 –have convinced many that astrophysical cosmology provides a new window on fundamentalphysics. These advances were possible because the observed structure takes the form oflinear perturbations of a simple state, an infinite uniform mixture of a small number ofcomponents with well understood interactions. The evolution of this system can be treatedrigorously and precisely. In addition, foreground effects are providentially weak and so causeonly minor complications. Fundamentalist physicists, drawn to cosmology by this success,often fail to appreciate the uniqueness of the circumstances. An interesting comparison ishelioseismology, the study of the structure of the Sun based on sound waves propagatingthrough it. Here also the perturbations are linear and propagate in a medium where therelevant physics is fully understood. Here also careful measurement has produced extremelyprecise results for the properties of a very large number of modes. Conclusions at thelevel of confidence and precision reached by CMB studies are precluded, however, by thecomplexity of the underlying system. For example, the initial fraction of heavy elementsrequired for the current standard model of solar evolution to reproduce the structure inferredfrom helioseismology is almost twice the fraction measured in the Sun’s atmosphere byanalysis of its spectrum.Astrophysical routes to a better understanding of Dark Energy all involve complexsystems: supernovae to trace the cosmic expansion history; galaxies to outline ripples inthe large-scale matter distribution; galaxies as background sources to trace gravitationallensing by the foreground mass distribution; galaxy clusters as markers of the growth ofcosmic structure. Astrophysical experience suggests that the ultimate precision reached bysuch programmes will be set by systematic effects, for example, progenitor or environmentevolution for supernovae, nonlinear and non-determinate relations between observables andtheoretical quantities for galaxies and galaxy clusters. By their nature such systematicscannot be accurately assessed in advance, and indeed they often remain unrecognised untilthe programme is complete. Estimates of the final precision of Dark Energy experiments arethus based primarily on purely statistical considerations and should be considered optimisticestimates of the ”best possible” result. Dark Energy enthusiasts, emboldened by CMBsuccesses, often fail to appreciate these limitations, believing that sophisticated statisticalanalysis will enable the best possible result to be approached. This exposes the communityto the danger of designing and carrying out a very expensive experiment to measure manythousand supernovae, or to image a very large area of sky, only to find that the resultingmeasurement of w is only a modest improvement over previous work because of astrophysicalsystematics. If the experiment is of limited use for other astrophysical purposes, then thefunds will, in effect, have been wasted. A problem for which the astrophysicists will surelybe blamed! 11 –The potential problem here reflects the combination of inappropriate risk assessment –what is the chance that the complexities of real galaxies, clusters or supernovae will frustrateattempts to measure the cosmic expansion and fluctuation growth histories precisely? –with an inappropriate design strategy – planning an experiment like WMAP rather than anobservatory like HST.This brings me to the second danger: the impoverishment of astrophysics by too heavyan emphasis on Dark Energy when planning the next generation of major facilities. Asalready discussed, astronomers traditionally limit risk when designing new instruments byconcentrating on the expansion of technical capabilities in sensitivity, wavelength coverage,spatial or spectral resolution. This enables progress on a wide variety of problems, partic-ularly since operation in observatory mode allows new projects to be proposed as they areseen to be interesting. Some Dark Energy projects conform to this strategy. For example,wide-angle X-ray and millimeter surveys will not only identify very large samples of distantgalaxy clusters, but will also image much of the sky to a sensitivity and resolution which hasnot previously been achieved at these wavelengths. In addition, these facilities will proba-bly operate at least to a limited extent in observatory mode. Other Dark Energy projects,for example those searching for supernovae or looking to measure baryonic features in thelarge-scale galaxy and mass distributions, will not extend previous sensitivity, resolutionor wavelength limits. Rather they achieve the required precision by observing much largerareas of sky than has previously been possible. Such surveys may not enable significantprogress in other areas of astrophysics. For example, deep photometric imaging of 2 squaredegrees of the sky has already been completed and provides data for hundreds of thousandsof faint galaxies. Rather few studies of the formation and evolution of galaxies would benefitfrom the 1000 times larger but otherwise similar samples provided by Dark Energy surveys.Since existing instrumentation can match the capabilities of these surveys, there is also littleincentive to operate them in observatory mode.The potential danger here is evident. The convergence between astronomers and funda-mental physicists produces a powerful lobby in favour of Dark Energy experiments. In thenatural competition between proposed large projects this works to the disadvantage of moretraditional observatories at X-ray, radio, ultraviolet or infrared wavelengths. These com-mand strong support only from astronomers, and so may be delayed, perhaps indefinitely,by financial constraints resulting from implementation of “higher priority” Dark Energy ex-periments. Astronomers will spend their time, energy and resources on experiments whichhave little impact on their main areas of research, while sacrificing the facilities which havetraditionally driven creativity, innovation and the advance of knowledge in their field. 12 –This leads to the third, and in my view most serious danger. By accepting the funda-mentalist view that Dark Energy is so important that clarifying its nature is the overidingproblem for current astrophysics, astrophysicists betray the underlying culture of their fieldand undermine its attractiveness both to future generations of creative scientists and to thepublic at large. This is exacerbated by other sociological trends within astrophysics which Inow digress briefly to discuss.In figure 3 I show bibliographic statistics, compiled from NASA’s Astrophysics DataSystem (ADS), to illustrate changes in astrophysics and space science over the last 30 years.In 1975 about 8500 different authors published a total of about 8900 papers in the refereedprofessional literature. By 2006 the number of authors had quadrupled but the number ofpapers had only doubled. On the other hand, the mean number of authors per paper alsodoubled, so that the number of papers signed by a typical astronomer remained constant atabout 2 per year. The size of the astronomical community has thus increased dramaticallyand a drop in the mean productivity of its members has been masked by the tendency formore individuals to sign each paper. In 1975 over 40% of all papers in the major journalshad a single author and fewer than 3% had 6 or more authors. In 2006 only 9% of papershad a single author while almost 28% had 6 or more authors. This trend towards team-basedprojects is undoubtedly real, but it is accentuated by the use of citations as a measure ofperformance, a practice which may influence another strong trend visible in figure 3: thereference lists of refereed astrophysics papers increased in length by a factor of 3.4 on averagebetween 1975 and 2006. Since the number of citations to individuals for a given year is theproduct of the number of papers, the mean length of their reference lists, and the meannumber of authors for the referenced papers, it was clearly much easier to get cited in 2006than in 1975 or even in 1995! As an extreme example, the fourth ranked astrophysicist bycitations to papers published over the last decade has never written a first-author paper for arefereed journal and has gained almost all his citations through his right to sign official papersby a large collaboration in which he played a purely functional role. The increasing numberof such survey collaborations, usually put together to justify large time investments on majorfacilities, means that more and more astrophysicists work in directed, quasi-functional roles,and that fewer achieve visibility through truly creative science.The concentration on large long-term projects has long dominated accelerator physics.Dark Energy projects will further accelerate this trend in astrophysics. Only with verylarge surveys can one hope for a percent level specification of the cosmic expansion andstructural growth histories. Achievement of these primary survey goals will have little impacton astrophysics beyond the Dark Energy issue, and most survey researchers will need toconcentrate on functional tasks to assure adequate data quality and timely completion of 13 –the project. Contrast this with the traditional, opportunistic style of the best astronomicalresearch, where individuals or small groups think up new ideas or build new instruments andapply them to situations where the scientific return seems likely to be greatest. A forward-looking observatory development programme can ensure that there are always new problemsto address and new opportunities to extend the scientific frontier. This is an attractive modelfor young researchers. They can have a major scientific impact already as graduate studentsand there is a clear path for them to establish themselves rapidly as independent playersin an international and exciting field. Such opportunities are rare in big survey science,particularly in many Dark Energy projects.This then is the third problem. If assembly of the very large surveys needed to constrainDark Energy comes to dominate astronomical research, then the development of other newcapabilities will be slowed, and opportunities to carry out creative individual research inmost areas of astrophysics will be reduced. This will make our subject less attractive tothe best and most ambitious young scientists, who will look to make their mark in otherdomains, biophysics or nanotechnology perhaps. Concentration on a single “fundamental”issue rather on the traditional diversity of issues will also make astronomy less attractiveto the general public, undermining taxpayer support for the expensive facilities needed topursue our science. Listening to the siren call of the fundamentalists may lose us both thecreative brains and the instruments that are needed to remain vibrant. Dark Energy is thePied Piper’s pipe, luring astronomers away from their home territory to follow high-energyphysicists down the path to professional extinction. What is to be done?
None of the negative consequences I have just outlined need necessarily follow from ourcurrent situation. My intention in this essay has been to draw attention to the dangers ofuncritically accepting that astronomers should spend much of their energy and resourcestrying to clarify the nature of Dark Energy, just because it is perceived as a fundamental(perhaps the fundamental) problem by high-energy physicists. In my view a hard-nosedcost-benefit analysis is needed, recognising both the inherent limitations of observationalastrophysics and the substantial cultural differences between the astronomy and high-energyphysics communities.Dark Energy is a deep and interesting puzzle which can be probed by astronomicalobservations alone, but it is one interesting puzzle among many and it may be one of the 14 –least likely to be “solved”. We do not know if astronomers can deliver measurements ofthe hoped-for precision, but even if they do, it seems likely that high-energy theorists willconstruct many Dark Energy models consistent with the observed expansion and structuregrowth histories. Dark Energy will be constrained, many possibilities will be excluded, butmany others will remain. Astronomers must be aware of this and must balance the needsof Dark Energy projects against those of the core areas of our field. New observatoriespromote exploration throughout astrophysics. They nourish the diversity and provide theopportunities for individual creativity which underlie its current flourishing. We must not beseduced by the fundamental nature of Dark Energy (and by the availability of new fundingsources) into sacrificing the foundation of our subject’s strength.Here are some suggestions for accepting Dark Energy as a prime subject for astronomicalstudy while embracing neither the fundamentalist view that it is the most important problemof our time, nor the industrial work patterns engendered by “Big Science” surveys of thekind required to significantly tighten constraints on its properties.1. Astrophysicists should recognise the cultural differences between their own field andhigh-energy physics. They should be willing to argue that astronomical discoveries –that the Universe expands, that the chemical elements were built in stars, that blackholes exist and can be far more massive than the Sun, that galaxies continually changeform, that other planets orbit other stars – although qualitatively different, are noless significant for humanity than the clarification of the underlying nature of forcesand particles. They should resist the fundamentalist argument that searching for theultimate structure of space, time and matter is deeper and more basic, and thus takesintellectual priority over other ways of extending our knowledge of the physical world.2. Large astronomical projects, even those for which Dark Energy issues are a primescience driver, should continue to be designed to push back the frontiers in many areasof astrophysics. Supernova surveys should store enough information to explore thesupernova mechanism and the relation of supernovae to the stellar populations fromwhich they form, as well to trace out the expansion history of the Universe. Galaxyredshift surveys should take sufficiently good spectra for a sufficiently well-definedset of galaxies that galaxy evolution can be studied, in addition to measuring thecharacteristic scale of galaxy clustering for use as a standard measuring rod. This issimply good astronomical practice, spreading the risk to compensate for the fact thatastronomers cannot ensure “proper” laboratory conditions for their experiments.3. Prioritisation of projects should be based not only on the case for their prime sciencegoal, but also on the extent to which they will enable future advances in astrophysics 15 –as a whole. In the case of Dark Energy surveys, this means recognising that refinementof the principal quantities to be measured, the cosmic expansion and linear fluctuationgrowth histories, is unlikely to impact significantly on other areas of physical cosmology.Thus, the enabling aspect of such surveys will come mainly from other science.4. Large projects require large teams and long time-scales. The negative effects of this onyoung scientists’ opportunities for creativity can be drastic and must be mitigated bypromoting a diverse set of science goals for exploration by young team members. Boththe Two Degree Field Galaxy Redshift Survey (2dFGRS) and the Sloan Digital SkySurvey (SDSS) were originally set up with a relatively narrow set of primary sciencegoals, but the teams involved were eventually able to address a very broad range ofproblems with their survey data, and many of these efforts were led by the youngerscientists. In the case of SDSS, the release of the full survey data through a powerful,publicly accessible database has allowed astronomers across the world to carry outtheir own SDSS projects, thereby enhancing the whole community’s opportunities forindividual creativity.5. Credit for scientific contributions must be clearly assigned to those responsible forthe original insights and for the creative aspects of the enabling work. Hard workalone brings little progress, and appropriate recognition is a prime incentive attractingcreative scientists to our field. Current assessment culture in astrophysics is basedmainly on total citations to papers signed by a scientist, regardless of whether (s)heis sole author or author number 47 out of 165. This encourages inflated author andreference lists which dilute the visibility of creative work over and above the dilutionalready caused by the trend towards large teams. This could be off-set in part bygreater reliance on first author citations (in astrophysics the first author is usually theperson with primary scientific responsibility for a paper) and on normalised citations(where an author is credited with 1 /N for a citation to a N -author paper). This wouldremove the temptation to inflate author lists and provide a fairer comparison of theoverall impact of individual astrophysicists. Unless we recognise them properly, thosecapable of original and creative contributions will prefer other fields.6. Astrophysicists should motivate their activities in their own cultural context, not inthat of high-energy physics. This is particularly important when interacting withstudents and young scientists. Dark Energy is undeniably a fascinating puzzle, but itis a high-energy physics puzzle. The creativity in understanding Dark Energy will notcome from planned astronomical surveys. They will provide more precise measurementsof quantities that are already well enough known for astrophysics. Although reachingsuch precision is a major challenge, it is a challenge that offers little opportunity for 16 –scientific creativity unless one is primarily interested in the processing of large datasetsor the statistics of data analysis. Bright and ambitious students will decide to becomeastrophysicists only if they see an opportunity to make high-impact contributions asindividuals. Within Dark Energy surveys, such opportunities will come mainly fromstudies of astronomical objects. In the rush to gain a funding edge by giving projects aDark Energy label, it is essential to avoid giving the impression that the astronomicalscience is “secondary”, of less significance or interest than improved measurementsof the cosmic expansion and structure growth histories. Indeed, it could turn outthat Dark Energy is more complex (or different) than most models suppose, and thatcritical clues to its nature emerge from traditional astronomical exploration of thephenomenology of structure, rather than from these precision measurements.Astronomy often claims to be the oldest of the physical sciences and it has a broadercultural and intellectual resonance with educated society than any other branch of physics.For this reason many university departments see astronomy as an ambassador for physics,providing the non-scientific public with some understanding of the scientific method anddrawing students into physics from a wide catchment area. The attraction lies in astronomy’sdiversity, in its combination of a lack of direct application to human society with insightsinto the development not only of our own world, but also of the larger cosmos in which it isembedded. These strengths are different from and complementary to those of fundamentalphysics. The continued vitality of astrophysics does not depend on its ability to constrainthe Deep Truth underlying all reality, but rather on its ability to retain our own and ourpublic’s fascination with the many-facetted views it offers of the processes which shaped ourUniverse and of the objects which populate it.
Acknowledgements
This essay grew out of a talk given in summer 2006 in the director’s Blackboard Lunchat the Kavli Institute for Theoretical Physics. I’d like to thank the director, David Gross,for the invitation to give a talk and for his spirited debate of the talk he got. I’d also like toacknowledge the unique atmosphere of this institute, which strives with remarkable successto promote cross-fertilisation between all branches of physics, fundamental and otherwise.Finally, I would like to thank Alberto Accomazzi of the Smithsonian/NASA AstronomicalData System for his help in compiling the publication statistics shown in figure 3. This debate can be viewed “live” at http://online.kitp.ucsb.edu/online/bblunch/white1/
17 –Fig. 1.— Two emblematic pictures from the Hubble Space Telescope. On the left is animage of the Eagle Nebula, a set of gas clouds illuminated by young stars and enshroudinga number of stars in formation. On the right is an image of the Hubble Deep Field. At thetime it was released in 1996 this was by far the deepest image of the sky ever made, showinggalaxies so distant that they are seen when the Universe was a small fraction of its presentage. 18 –Fig. 2.— Two emblematic pictures from the Wilkinson Microwave Anisotropy Probe. Atthe top is the WMAP map of temperature fluctuations in the cosmic microwave backgroundradiation. These fluctuations are very weak, with typical amplitudes of a few parts in100,000. They are a direct image of structure in the Universe when it was only 400,000years old. Below the map, its power spectrum (the points with error bars) is comparedwith an a priori model (the smooth curve) which assumes that all structure originated asquantum zero-point fluctuations during a very early period of inflationary expansion. Thesix parameters specifying the model all have physical meanings and they all take valueswhich are quite compatible with those inferred from independent astronomical observationsof the nearby Universe. 19 –1980 1990 20001234 DateRefereed papers (8503)References/paper (9.17)Authors/paper (2.08)Distinct Authors (8916)