ASTROBIOLOGY

Volume 16, Number 5, 2016 Research Article

ª Mary Ann Liebert, Inc.
DOI: 10.1089/ast.2015.1418

A New Empirical Constraint on the Prevalence

of Technological Species in the Universe

A. Frank1 and W.T. Sullivan III2

Abstract

In this article, we address the cosmic frequency of technological species. Recent advances in exoplanet studies
provide strong constraints on all astrophysical terms in the Drake equation. Using these and modifying the
form and intent of the Drake equation, we set a firm lower bound on the probability that one or more
technological species have evolved anywhere and at any time in the history of the observable Universe. We
find that as long as the probability that a habitable zone planet develops a technological species is larger than
*10-24, humanity is not the only time technological intelligence has evolved. This constraint has important
scientific and philosophical consequences. Key Words: Life—Intelligence—Extraterrestrial life. Astrobiology
2016, xxx–xxx.

1. Introduction 2. Method

T he history of physics has shown that fundamental in-

sights into a problem can sometimes be acquired by
setting order-of-magnitude estimates of scales, limits, and
Our approach asks a very different question from the
usual treatment of the subject. Standard astrobiological
discussions of intelligent life focus on how many techno-
boundaries via combinations of constants or parameters. This logical species currently exist with which we might com-
approach can be particularly effective for problems in which municate (Vakoch and Dowd, 2015). But rather than asking
few empirically derived constraints are available, such as the whether we are now alone, we ask whether we are the only
derivation of the quantum gravitational Planck length, that is, technological species that has ever arisen. Such an approach
lp = (hG/c3)1/2, or early studies using the cosmic density pa- allows us to set limits on what might be called the ‘‘cosmic
rameter (O = r/rc). Here we adopt such an approach to ad- archaeological question’’: How often in the history of the
dress an overarching goal of astrobiology: understanding if Universe has evolution ever led to a technological species,
humanity is ‘‘alone’’ (Sullivan and Baross, 2007; Impey whether short- or long-lived? As we shall show, providing
et al., 2012). One way of framing this question is to ask if constraints on an answer to this question has profound
other technology-building species exist now, or have ever philosophical and practical implications.
existed at any time, in the Universe. Given that we have only We first modify the Drake equation in order to address
one known example of a planet where such evolution has how many technological species have formed over the his-
occurred (and also only one example of where even microbial tory of the observable Universe. We call this number A (for
life has evolved), there seems to be little hope in determining archaeology) and use it to investigate the probability that
the fecundity of the Universe in producing technological humanity is unique (i.e., A = 1). Note that we are explicitly
species. This conclusion, however, ignores the rapid and not concerned with the average lifetime CLD of such species
substantial progress astrobiology has made in the last two or if they still exist such that we could receive their signals
decades. In particular, the empirical determination of exo- or signal them. This is in contrast to the usual Drake
planet statistics has radically changed the nature and quality equation formulation, which calculates Nc, the number of
of constraints astrobiologists now have at their disposal when technological species now existing (hence its concern with
considering the prevalence of life in the Universe. In this CLD). Given this approach, effects of time are removed where
article, we employ these new constraints to set a lower limit they normally appear in the form of star-formation rates,
on the probability that technological species have ever stellar lifetimes, technological species lifetimes, and distri-
evolved anywhere other than on Earth. bution of epochs of arising. As usual for the Drake equation,

1
Department of Physics and Astronomy, University of Rochester, Rochester, New York.
2
Department of Astronomy and Astrobiology Program, University of Washington, Seattle, Washington.

1
2 FRANK AND SULLIVAN

we assume that technology is associated with planets and suring that Earth is the only location in the history of the
their host stars. cosmos where a technological civilization has ever evolved.
We define the ‘‘A-form’’ of the Drake equation, which Adopting A = 0.01 means that in a statistical sense were we
describes the total number of technological species that have to rerun the history of the Universe 100 times, only once
ever evolved anywhere in the currently observable Uni- would a lone technological species occur. A lower bound fbt
verse: on the probability is then

A ¼ [N fp np ][fl fi ft ] (1) fbt ¼ A=Nast ¼ 0:01=4 · 1021 ¼ 2:5 · 10  24

¼ Nast fbt
Thus only for values of the product fbt lower than
where N* is the total number of stars, fp is the fraction of 2.5 · 10-24 are we likely to be alone and singular in the
those stars that form planets, np is the average number of history of the observable Universe. This limiting value of fbt
planets in the habitable zone of a star with planets, fl is the can be considered to define a ‘‘pessimism line’’ in discus-
probability that a habitable zone planet develops life, fi is the sions of the prevalence of technological civilizations on a
probability that a planet with life develops intelligence, and cosmic scale. On the other hand, if evolutionary processes
ft is the probability that a planet with intelligent life devel- lead to higher values, we can be assured that we are not the
ops technology (of the ‘‘energy intensive’’ kind such as that only instance in which the Universe has hosted a techno-
of our own civilization). logical species.
The second version of Eq. 1 reduces the right-hand side to We can generalize these results for structural scales (of
two factors, the first of which, Nast, includes all factors in- size Rs) within the Universe: galaxies, clusters of galaxies,
volving astrophysics and represents the total number of and superclusters of galaxies (Fukugita and Peebles, 2004).
habitable zone planets. The second factor, fbt, gathers the Table 1 provides typical values for these entities and lists
three factors involved with biology, evolution, and ‘‘planetary the corresponding values of fbt for each of them, as well as
sociology,’’ and represents the total ‘‘biotechnical’’ proba- for the entire Universe. For instance, for our own galaxy fbt
bility that a given habitable zone planet has ever evolved a is 1.7 · 10-11, meaning that another technical species has
technological species. The factor fbt = flfift is extremely un- likely occurred in the history of the Milky Way if the
certain because (a) we have no theory to guide any estimates probability of a technological species arising on a given
and (b) we have only one known example of the occurrence planet in a habitable zone is greater than one in 60 billion.
and history of life, intelligence, and technology. In what Figure 1 presents these results for given values of fbt. The
follows, we leave fbt as statistically unknown at this time and figure allows one to see the corresponding number of
examine the consequences of its taking on various values technological species that have ever arisen on various scales
depending on one’s pessimism or optimism. for various assumptions of the difficulty of biology and
The key development that enables our new approach is technology to evolve.
that observations now provide accepted or statistical well-
determined values for all factors contributing to Nast. The 4. Discussion and Conclusions
importance of this accomplishment cannot be overstated—
These results have wide implications. First, the long his-
until recently only one factor (N*) was known, and it was
tory of debate over extraterrestrial intelligence can be char-
entirely possible that habitable zone planets might have
acterized as one of pessimists versus optimists relative to
been extremely rare (Nast/N* << 1). But the combination of
choices for the various factors in the Drake equation (Vakoch
radial velocity, transit, and microlensing based methods
and Dowd, 2015). These debates have, however, been un-
now yields statistically well-constrained values for both fp
constrained in terms of how pessimistic one can be. With our
and np.1
approach we have, for the first time, provided a quantitative
We adopt the values fp * 1.0 (Cassan et al., 2012) and np
and empirically constrained limit on what it means to be
* 0.2 (Petigura et al., 2013); therefore Nast/N* = 0.2. Note
pessimistic about the likelihood of another technological
that the number of stars N* is also an observable quantity,
species ever having arisen in the history of the Universe. We
which depends on the size of the region being considered.
have done so by segregating newly measured astrophysical
factors from the fully unconstrained biotechnical ones, and by
3. Results
shifting the focus toward a question of ‘‘cosmic archaeology’’
We now turn to the specific question, ‘‘Has even one and away from technological species lifetimes. Our constraint
other technological species ever existed in the observable addresses an issue that is of particular scientific and philo-
Universe?’’ We take N* = 2 · 1022 for the total number of sophical consequence: the question ‘‘Have they ever ex-
stars in the observable Universe (Silburt et al., 2015) To isted?’’ rather than the usual narrower concern of the Drake
address our question, A is set to a conservative value en- equation, ‘‘Do they exist now?’’ Perhaps in the long term we
should contemplate undertaking a field survey in cosmic ar-
chaeology, seeking possible evidence for such past technical
1
Rather than fp and np, some authors focus on ZE, defined as the species (Stevens et al., 2015).
occurrence rate of Earth-sized planets (1–2 Earth radii) in the Secondly, we note that sample sizes A of order 100 or
habitable zones of Sun-like stars. For example, one recent study 1000 have a particular importance in current discussions
based on Kepler mission transiting planets finds a value ZE * 0.06
(Silburt et al., 2015). For the purposes of our order-of-magnitude
over efforts to create a sustainable, energy-intensive, high-
estimates, which also include non-Sun-like stars, we stay with the technology civilization here on Earth (Frank and Sullivan,
factors in the conventional Drake equation. 2014). Given the challenges human society faces from
A CONSTRAINT ON TECHNOLOGICAL SPECIES IN THE UNIVERSE 3

Table 1. Limits on the Number of Technological Species Occurring on Different Scale Lengths
Scale length Size Rs (M lt-yr) No. galaxies Nast fbt for A = 1
Galaxy 0.1 1 6 · 10 10
1.7 · 10-11
Galaxy cluster 5 300 2 · 1013 5 · 10-14
Supercluster 300 3000 2 · 1014 5 · 10-15
Observable Universe 13,700 7 · 1010 4 · 1021 2.5 · 10-22

climate change, resource allocation, and loss of species di- changes, hydrological cycle disruptions, etc.; Frank and
versity, it is not clear if the kind of long-term global culture Sullivan, 2014). Thus modeling ensembles of such species
we hope to build is even sustainable. However A > 100 in to understand the broad classifications of their planetary
our formulation implies that the evolution of technological feedback histories is theoretically well-grounded for large-
species has occurred enough times that the ensemble of their enough values of fbt. Such a project would be an attempt to
histories (or trajectories in a suitably defined phase space) is understand both average properties like CLD as well as what
statistically meaningful. In particular, for large-enough led some trajectories to collapse and others to long-term
values of fbt, the average longevity CLD of such a sample of sustainability.
technological species does, in principle, exist. Thus if fbt is It is also possible to make the connection between our A
such that A ‡ 1000 (the statistically relevant population), it is form of the Drake equation and traditional SETI studies that
reasonable to consider that different versions of humanity’s ask about Nc, the number of technological species existing
current technological experiment in the alteration of our in our galaxy now. The traditional form of the Drake
planetary system have occurred before. Note that many as- equation can be written
pects of the feedback between an energy-intensive, tech-
nological species and its host planet would depend solely on ÆLæ
physical processes and constraints (atmospheric chemistry Nc ¼ ½Nast ½fbt 
L

Thus Nc ¼ A ÆLæ
L . As we have shown in Table 1, for each
astrophysical scale of interest s, we candefine a minimum
biotechnical probability via Asmin ¼ Nast
s s
fbt, min ¼ 1. Using
this, we can then ask at each scale how much does fbts have to
s s
be increased above fbt, min to get some desired value A
s
above Amin . Thus we can write

 s  s  ÆLæ
Ncs ¼ Nast fbt, min As
L

Taking L*, the lifetime of a star, to be *1010 yr and using

Table 1 for a single galaxy (our own), one has

   ÆLæ
Ncg ¼ 6 · 1010 1:7 · 10  11 As ¼ 10  10 As ÆLæ
1010 yr

In order for there to be one other partner technological

species in the Galaxy with which to make contact (Ncg ¼ 2),
the product AsCLD must be equal to 2 · 1010 yr. For an op-
timistic upper limit of CLD = 106 yr, As must be at least
2 · 104 (i.e., there must have been at least 20,000 techno-
logical species spawned across the lifetime of the Galaxy).
For a perhaps more realistic guess (based on our own civi-
FIG. 1. The number A of technological species that have lization) of CLD = 104 yr, As = 2 · 106.
ever occurred over the history of the Universe versus the Casting these results in a different way, one can imagine
scale size Rs of different hierarchical structures in which conducting a general Milky Way SETI search that might
their host stars are found: galaxies, clusters of galaxies, hope for one in every million stars to host a technical spe-
superclusters of galaxies, and the entire Universe. Reference cies today, which means Ncg ¼ 3 · 105 . To achieve this
values of A = 0.01, 1, and 100 are marked. The four curves number, if we very optimistically adopt fbt = 1, then CLD must
represent different values of fbt, the probability of a tech-
nological species arising on a given habitable zone planet. be 5 · 104 yr. More realistic lower values of fbt require
Only if fbt falls as low as 2.5 · 10-24 to 2.5 · 10-22 is it likely proportionately higher values for CLD to reach the desired
that no other technological species has ever arisen in the number of potential finds. Note that in all cases the number
entire Universe. Color images available online at www of technological species that have ever existed in our galaxy
.liebertpub.com/ast would be 6 · 1010 fbt.
4 FRANK AND SULLIVAN

In conclusion, we have shown that recent advances in Fukugita, M. and Peebles, P.J.E. (2004) The cosmic energy
astrobiology and exoplanet studies mean that an empirically inventory. Astrophys J 616:643–668.
derived lower limit can now be placed on the probabilities Impey, C., Lunine, J., and Funes, J., editors. (2012) Frontiers of
that even one other technological species has ever evolved Astrobiology, Cambridge University Press, Cambridge, UK.
in the Universe or in our galaxy. This limit provides a Petigura, E.A., Howard, A.W., and Marcy, G.W. (2013) Pre-
framework for discussions of both life in its cosmic context valence of Earth-size planets orbiting Sun-like stars. Proc
and questions about trajectories of technological species Natl Acad Sci USA 110:19273–19278.
relevant to our own issues of global sustainability. Silburt, A., Gaidos, E., and Wu, Y. (2015) Statistical recon-
struction of the planet population around Kepler solar-type
Acknowledgments
stars. Astrophys J 799, doi:10.1088/0004-637X/799/2/180.
Stevens, A., Forgan, D., and O’Malley James, J. (2015) Ob-
We thank Dan Watson and David Catling for comments servational signatures of self-destructive civilisations. arXiv
on the text, and Matt McQuinn, Caleb Scharf, and Gavin 1507.08530
Schmidt for discussion. Sullivan, W.T. and Baross, J.A., editors. (2007) Planets and
Life: The Emerging Science of Astrobiology, Cambridge
References University Press, Cambridge, UK.
Vakoch, D.A. and Dowd, M.F., editors. (2015) The Drake
Cassan, A., Kubas, D., Beaulieu, J.-P., Dominik, M., Horne, K., Equation: Estimating the Prevalence of Extraterrestrial
Greenhill, J., Wambsganss, J., Menzies, J., Williams, A., Jør- Life through the Ages, Cambridge University Press, Cam-
gensen, U.G., Udalski, A., Bennett, D.P., Albrow, M.D., Ba- bridge, UK.
tista, V., Brillant, S., Caldwell, J.A.R., Cole, A., Coutures, Ch.,
Cook, K.H., Dieters, S., Prester, D.D., Donatowicz, J., Fouqué, Address correspondence to:
P., Hill, K., Kains, N., Kane, S., Marquette, J.-B., Martin, R., A. Frank
Pollard, K.R., Sahu, K.C., Vinter, C., Warren, D., Watson, B., Department of Physics and Astronomy
Zub, M., Sumi, T., Szymański, M.K., Kubiak, M., Poleski, R., University of Rochester
Soszynski, I., Ulaczyk, K., Pietrzyński, G., and Wyrzykowski, Rochester, NY 14620
q. (2012) One or more bound planets per Milky Way star from
microlensing observations. Nature 481:167–169. E-mail: [email protected]
Frank, A. and Sullivan, W. (2014) Sustainability and the as-
trobiological perspective: framing human futures in a plane- Submitted 5 October 2015
tary context. Anthropocene 5:32–41. Accepted 16 February 2016