Inspired by a discussion at the Technoclimes workshop, I started thinking about all of the different approaches to SETI, as distinct from the different technosignatures to search for. This eventually evolved into a paper where I was able to incorporate lots of odds and ends I had written and collected over the years about SETI in one place.  I think it came out well!  It’s on the arXiv here, but here are some of the highlights.

There are a few ways to think about SETI searches, and most fall onto one side of a few divides:

1. Communication vs. Artifacts, including:
   • Small vs. large scale
   • Kinds of artifacts / carriers
   • Derelict vs. active artifacts
2. Ambiguous vs. Dispositive Technosignatures
3. Commensal/Archival vs. Dedicated Searches
4. Model-based vs. Anomaly Searches
5. Searching for “Beacons” vs. Eavesdropping
6. Passive vs. Active Searches (i.e. METI).

Communication vs. Artifacts:

Last year we put a lot of effort into proposing a big interdisciplinary research consortium for astrobiology to NASA (an “ICAR”) . The proposal was unsuccessful, but along the way we found a useful way to frame searches:

The idea here was to map out the kinds of “artifact” technosignatures that exist, and think about how they all relate.  Along the top roughly tracks both scale and distance: first nearby things in the solar system; next, roughly Type I Kardashev scale things on the surfaces or in orbit around nearby planets; then things approaching Type II with lots of circumstellar activity; and finally Type II sorts of technosignatures on the right.

Vertically, we take three things to actually look for: physical structures, environmental alteration, and excess heat.  There are many other kinds of technosignatures too, of course. In particular, communication SETI is not really on this chart, but  by and large I think this captures the breadth of a pretty big swath of SETI.

Artifacts are neat because we might be able to detect them even if they are no longer being maintained. Depending on the artifact, they might be detectable for a very long time after their creators are gone.

Ambiguous vs. Dispositive Technosignatures

“Dispositive” means that something settles (“disposes of”) a particular question. It’s a term from law, and I like it because it’s a useful word with no synonyms and can help distinguish among different kinds of null results (i.e. failing to find anything because you didn’t look hard enough, and showing something does not exist because you looked more than enough).

One of the really nice things about communication SETI is that it’s probably dispositive: if you see a communicative signal, especially a narrowband one, you know it’s from technology. Then you’ve solved several problems at once, scaling the entire “Ladder of Life Detection” in one go.

Hunts for Dyson Spheres, on the other hand, and not very dispositive. Waste heat can come from dust just as well as from technology, and no matter how weirdly shaped a light curve implies an occulting object is, there always seems to be some pathological natural explanation for it.  Such searches can, at best, find good candidates for technosignatures that would then have to be validated by other means.

But, that’s also true of many searches for biosignatures! Just finding, say, oxygen isn’t enough. Hey, no one said astrobiology was going to be easy!

Commensal/Archival vs. Dedicated Searches

Some kinds of searching requires dedicated hardware. The Breakthrough Listen Initiative builds large supercomputers on site at their facilities to record voltages measured at the telescopes extremely quickly and save the reduced data products. PanoSETI will perform unprecedented observations of the transient sky because of its innovative design.

Other searches are “commensal,” using some hardware simultaneously with other observers. The old SERENDIP project at Arecibo had specialized hardware that occupied a different part of the focal plane from the main instruments, and so was always on, searches whatever part of the sky it could. This kind of searching sacrifices the ability to choose one’s targets for getting in order to use powerful equipment.

Other projects can be done with no additional hardware. Searches that use archival data, for instance, piggyback on general purpose astronomy data.  This kind of searching is cheap, but one can only search for things that happen to fall within the parameters of the databases you’re looking at.

Model-based vs. Anomaly Searches

This is a neat one. There’s an idea called “generalized SETI” by George Djorgovski which roughly says that since we don’t really know what form alien technology will take, you should look for anything out of the ordinary in big public data sets.  This has the upside that you are sensitive to the unexpected, and the idea has been applied to many aspects of SETI, for instance this work by Daniel Giles.

Model-based searches look for a particular technosignature in that data by modeling its signal and filtering on that. This has the big advantage that you can say what it is you did not find, because you have parameterized your technosignature and its strength. This makes it much easier to calculate upper limits.

Putting such upper limits on anomaly-based searches can be much harder because it can be difficult to know, especially with machine learning algorithms, exactly what it is the computer is keying on, on what it would have missed. This is a major problem worth tackling, because anomaly-based searches have enormous promise.

Searching for “Beacons” vs. Eavesdropping

In the early days of SETI, a mix of optimism and necessity led Frank Drake and others to search for “beacons”—big, loud, obvious signals designed to get our attention. Perhaps, one line of reasoning went, there was a community of species welcoming technologically young species like ours into their Galactic Society with such signals.

Such signals are easy to spot because they would be designed to be easy to spot, and so it makes sense to look for them first.  This was especially true because early radio equipment could search only a very narrow range of frequencies at once. A lot of work went into thinking about what frequencies such beacons would be at—the first SETI paper (Cocconi & Morrison 1959) guessed the 21-cm line, and many “magic frequencies” have since been proposed as the ones “they” would use to get our attention.

Indeed, there is a whole concept in game theory called “Shelling points” used to describe this dynamic.

Today, radio observatories can search billions of channels simultaneously, eliminating the need to guess, and our sensitivity is much better, so we could potentially detect even “leaked” emission intended only for short-range communication, for instance among planets in a distant planetary system. So the distinction is no longer quite so important, but still influences survey design and target selection.

Passive vs. Active Searches (i.e. METI)

This one gets a lot of attention!  METI is the attempt to establish contact backwards: to send a signal that gets attention in hopes that it triggers a response we wouldn’t miss. Some people get very upset about METI, worried we might catch the attention of dangerous aliens. My position is pretty nuanced, but ultimately I’m not worried about METI. Earth has many signatures of life and technology that I think are more obvious than any METI programs, so ultimately their value is performative, to get us thinking about contact.

Next time: Upper limits!