Given that we are just emerging as a spacefaring species, it seems reasonable to think that any civilizations we are able to detect will be considerably more advanced — in terms of technology, at least — than ourselves. But just how advanced can a civilization become before it does irreparable damage to itself and disappears? This question of longevity appears as a factor in the famous Drake Equation and continues to bedevil SETI speculation today.
In a paper in process at The Astronomical Journal, Amedeo Balbi (Università degli Studi di Roma “Tor Vergata”) and Milan Ćirković (Astronomical Observatory of Belgrade) explore the longevity question and create a technosignature classification scheme that takes it into account. Here we’re considering the kinds of civilization that might be detected and the most likely strategies for success in the technosignature hunt. The ambiguity in Drake’s factor L is embedded in its definition as the average length of a civilization’s communication phase.
Immediately we’re in shifting terrain, for in the early days of SETI, radio communication was the mode of choice, but even in the brief decades since Project Ozma, we’ve seen our own civilization drastically changing the radio signature it produces through new forms of connection. And as Balbi and Ćirković point out, the original L in Drake’s equation leaves open a rather significant matter: How do we treat the possibility of civilizations that have gone extinct?
These two authors have written before about what they call ‘temporal Copernicanism,’ which leads us to ask how the longevity of a civilization is affected by its location in our past or in our future. We are, after all, dealing with a galaxy undergoing relentless processes of astrophysical evolution. As we speculate, we have to question a value for L based on a civilization (our own) whose duration we cannot know. How can we know how far our own L extends into the future?
Image: Messier 107, a globular cluster around the disk of the Milky Way in the constellation Ophiuchus, is a reminder of the variety of stellar types and ages we find in our galaxy. What kind of technosignature might we be able to detect at a distance of about 20,000 light-years, and would ancient clusters like these in fact make reasonable targets for a search? Many factors go into our expectations as we formulate search strategies. This image was taken with the Wide Field Camera of Hubble’s Advanced Camera for Surveys. Credit: ESA/NASA.
Thinking about these matters always gets me thinking of Arthur C. Clarke’s 1956 novel The City and the Stars, set in the city of Diaspar a billion years from now. How do we wrap our heads around a civilization measured not just in millennia but in gigayears? Speculative as they are, I find a kind of magic in playing around with terms like Γ, cited here as the average rate of appearance of communicating civilizations (with L, as before, their average longevity), so that if we take Γ as constant in time, its value can be estimated as the total number of technosignatures over the history of the galaxy (Ntot) divided by the age of the galaxy (TG). Thus Balbi and Ćirković cite the equation:
From the paper:
It is apparent that the number of technosignatures that we can detect is a fraction of the total number that ever existed: the fraction is precisely L/TG. Because TG ∼ 1010 years, L/TG is presumed to be generally small; any specific precondition imposed on the origination of technosignatures, like the necessity of terrestrial planets for biological evolution, will act to reduce the fraction. This is the quantitative argument that justifies one of the most widely cited assertions of classical SETI, i.e. that the chances of finding ETIs depend on the average longevity of technological civilizations. (In fact, it is well-known that Frank Drake himself used to equate N to L.)
The equation clarifies the idea that SETI depends upon the average longevity of technological cultures, but the authors point out that another way to look at the matter is this: L needs to be large, for we’re requiring a high number of technosignatures indeed to have any chance for detecting a single one. Spread out over time, many such signatures need to have existed for us to make a single detection, or at best a few, with our present level of technology.
And here is where Balbi and Ćirković take us away from the more conventional approach derived above. Is the number of detectable technosignatures, N, static over time? From the paper:
…both Γ and L are average quantities, and there is an implicit assumption that N is stationary over the history of the Galaxy. There are good reasons to believe that this is not the case. Of course, it is unrealistic to assume that Γ is constant with cosmic time. Even if we limit ourselves to the last ≃ 10 Gyr of existence of thin disk Pop I stars which are likely to harbour the predominant fraction of all possible habitats for intelligent species, their rate of emergence is likely to be very nonuniform. One obvious source of nonuniformity is the changing rate of emergence of planetary habitats, as first established by Lineweaver (2001) and subsequently elaborated by Behroozi & Peeples (2015), as well as by Zackrisson et al. (2016). This nonuniformity can be precisely quantified today and some contemporary astrobiological numerical simulations have taken it into account (Ðošović et al. 2019).
We should assume, the authors argue, that the appearance of technosignatures varies with time. They are interested less in coming up with a figure for N — and again, this is defined in their terms (not Drake’s) as ‘the number of detectable technosignatures’ — than in spotlighting the most likely type of technosignature we can detect. Their classification scheme for technosignatures as filtered through the lens of longevity goes like this:
Type A: technosignatures that last for a duration comparable to the typical timescale of technological and cultural evolution on Earth, τ ∼ 103 years
Type B: technosignatures that last for for a duration comparable to the typical timescale of biological evolution of species on Earth, τ ∼ 106 years
Type C: technosignatures that last for for a duration comparable to the typical timescale of stellar and planetary evolution, τ ∼ 109 years
The scheme carries an interesting subtext: The longevity of technosignatures does not have to coincide with the longevity of the species that created the detectable technology. Here we’re at major variance from the L in Frank Drake’s equation, which had to do with the lifetime of a civilization that was capable of communicating. Balbi and Ćirković are tightly focused on the persistence not of civilizations but of artifacts. Notice that a technosignature search is likewise not limited to planetary systems — an interstellar probe could throw its own technosignature.
We might assume that technosignatures of long duration could only be produced by highly advanced civilizations capable of planetary engineering, say, but let’s not be too sure of ourselves on that score, for some technosignatures might be left behind by species well down on the Kardashev scale of civilizations. Consider Breakthrough Starshot, for example. Let’s push its ambitions back a bit and just say that perhaps within a century, we may be able to launch flocks of small sailcraft to nearby stars using some variation of its methods.
These would constitute a technosignature if detected by another civilization, as would remnant probes like Voyager and Pioneer, as would some forms of atmospheric pollution or simple space debris. A single civilization could readily produce different kinds of technosignatures over the course of its lifetime. As the authors note:
Our species has not yet produced Type A technosignatures, if we only consider the leakage of radio transmissions or the alteration of atmospheric composition by industrial activity; but its artifacts, such as the Voyager 1 and 2, Pioneer 10 and 11, and New Horizons probes, could in principle become type B or even C in the far future, even if our civilization should not survive that long. Similarly, a Type C technosignature can equally be produced by a very long-lived civilization, or by one that has gone extinct on a shorter time scale but has left behind persistent remnants, such as a beacon in a stable orbit or a Dyson-like megastructure.
Persistent remnants. I think of the battered, but more or less intact, Voyager 2 as it passes the red dwarf Ross 248 at about 111,000 AU some 40,000 years from now (Ross 248 will, in that era, be the closest star to the Sun). That’s a technosignature waiting to be found, one produced by a civilization low on the Kardashev scale, but it bears the same message, of a culture that explores space. I wonder what kind of a technosignature Clarke’s billion year old civilization in Diaspar would have thrown?
Whatever it might be, it would surely be more likely to be detected than our Voyager 2, a stray bit of flotsam among the stars. That said, I keep in mind what we learned from the TechnoClimes workshop — and Jim Benford’s continuing work on ‘artifact’ SETI — making the point that we can’t rule out a local artifact in our own system. And, of course, if Avi Loeb is correct, we may already have found one, though suitably ambiguous in its interpretation. Clearly, if we did detect technosignatures close to home, the implication would be that they are found widely in the galaxy, and that would dramatically change the nature of the hunt.
So the scope for technosignatures is wide, but drawing the lessons of this paper together, the authors find that the technosignature we are most likely to detect with present technological tools is a long-lived one, meaning in Balbi and Ćirković’s terms, one with a duration of at least 106 years. Technosignatures younger than this may be detectable but only if it turns out they are common, as thus relatively nearby and easier for us to find. Of course we can search for them, but the authors believe these searches are unlikely to pay off. Their thought:
This suggests that an anthropocentric approach to SETI is flawed: it is rational to expect that the kind of technosignatures we are most likely to get in contact with is wildly different, in terms of duration, from what has been produced over the course of human history. This conclusion strengthens the case for the hitherto downplayed hypothesis (which is not easily labeled as “optimistic” or “pessimistic”) that a significant fraction of detectable technosignatures in the Galaxy are products of extraterrestrial civilizations which are now extinct.
How to proceed? The authors’ focus on longevity leads them to conclude that our most likely targets may well be rare and they may flag extinct civilizations, but the value N that Balbi and Ćirković are talking about is different than classical SETI’s N, which needs a large value to ensure detection. It only takes one technosignature, and a few of the Type C signatures would be much more likely to be detected than a spectacularly high number of Type A signatures:
Dysonesque megastructures, interstellar probes, persistent beacons—as well as activities related to civilizations above Type 2 of the Kardashev scale, or to artificial intelligence—should be the preferred target for future searches. These technosignatures would not only be ‘weird’ when measured against our own bias, but could arguably be less common than short-lived ones. Such [a] conclusion deflates the emphasis on large N (and human-like technosignatures) that informed much of classical SETI’s literature.
If this sounds discouraging, it need not be. It simply tells us the kind of strategy that has the greatest chance for success:
…the supposed rarity of long-lived technosignatures should not be regarded, in itself, as a hindrance for the SETI enterprise: in fact, a few Type C technosignatures, over the course of the entire history of the Galaxy, would have much higher chance of being detected than a large number of Type A. Also, possible astrophysical mechanisms which could lead to a posteriori synchronization of shorter lived technosignatures should be investigated, to constrain the parameter space of this possibility, if nothing else.
Civilizations that appeared long ago and survived have conceivably found a way to persist, and therefore may still be active, but for detection purposes their existence now is less significant than what they may have left behind. Just how they grew to the point where they could begin the construction of detectable technosignatures is explored in the paper’s discussion of ‘phase-transition’ scenarios via a mathematical framework used to model longevity. “Achieving such form[s] of institutions and social structures might count as an advanced engineering feat in its own right,” as the authors note.
Technosignature work is young and constitutes a significant extension of the older SETI paradigm. Thus modeling how to proceed, as we saw both here and in the previous post on NASA’s TechnoClimes workshop, is the only path toward developing a search strategy that is both sound in its own right and also may have something to teach us about how our own civilization views its survival. The kinds of insight technosignature modeling could produce would take us well beyond the foolish notion of some early SETI critics that its only didactic function is as a form of religion, looking for salvation in the form of the gift of interstellar knowledge. To the contrary, the search may tell us much more about ourselves.
The paper is Balbi and Ćirković, “Longevity is the key factor in the search for technosignatures,” in process at The Astronomical Journal (preprint).