Percival Lowell and the Canals of Mars, Part II: How to See Things That Aren’t There

Matthew J. Sharps

Seth Hurd, Brandon Hoshiko, Erik Wilson, Mary Alyssa Flemming, Simran Nagra, and Maribell Garcia (co-authors) are research students in the Cognitive Science Laboratory under Sharps’s direction, Department of Psychology, California State University, Fresno.


The same errors of observation and interpretation that led Percival Lowell and others to see “canals” on Mars can happen today. New experiments show that 30 percent of presumably rational respondents provided completely imaginary accounts of structures while looking at a blank white disk.

Near the turn of the twentieth century, several distinguished astronomers reported the observation of “channels” or “canals” on Mars. These canals, as planetary features, don’t exist at all (e.g., Amundson 2017; Brasch 2018; Sheehan 1988). Rather, as discussed in a previous SI article (Mat 2018), the “observation” of these canals derived from psychological factors in the minds of the observers.

Our present research focuses on the ways in which these psychological factors can literally alter subjective reality to the point that people see things that aren’t there at all. Including canals. Making this happen is easier than one might think. 

One of the most important reporters of the canal phenomenon was the great planetary astronomer Percival Lowell. Like many responsible scientists of his day, Lowell saw the “canals” as artificial water channels designed by something like little green men with little green civil engineering degrees. Lowell thought that these alien hydrologists were trying to save their increasingly desiccated desert civilization by shipping water around in their extraterrestrial canals (see especially Amundson  2017; Brasch 2018); however, even in Lowell’s day—and even with this heroic if somewhat melodramatic alien backstory—many scientists did not find the canals convincing. 

In a paper published in 1903, Joseph Edward Evans and Edward Walter Maunder set out to prove that the martian canals were a mental rather than physical phenomenon. In a series of experiments, British schoolboys were asked to copy maps of Mars (with the major confirmed features of the planet already drawn on them). Evans and Maunder found that many of the boys connected major features with lines, which coincided with Lowell’s “canals”; these nonexistent linear structures had no possible genesis outside the boys’ own minds, exactly as suggested by the principles of Gestalt psychology (see King and Wertheimer 2005; Kohler 1947; Sheehan 1988; also see Sharps 1993; Sharps 2003; Sharps 2018).

Granted, there are psychological differences between schoolboys and adult astronomers that might limit the generalizability of this work. Also, a modern experimental psychologist tends to gasp in horror at some of the omissions in the published report of this research (for example, the report includes the phrase “There will be no need to give in detail all the experiments made … .” [Evans and Maunder 1903]). However, whatever the problems with methods and reportage, Evans and Maunder clearly demonstrated that there was a psychological source of “canals” quite outside the rather outlandish possibility of martian construction gangs energetically plowing through the alien topsoil with extraterrestrial steam shovels.

On the other hand, the loose methodology of the Evans and Maunder work may have had something to do with the later contradictory work of the astronomer Flammarion, who utterly failed to replicate these English results with French schoolboys. The French drawings were effectively bereft of canals (see Sheehan 1988). The entire canal thing remained, so to speak, up in the air.

There were other dissenters from Lowell’s perspective as well. Lowell’s assistant, A.E. Douglass, began to worry about the entire canal business when he found that he, Douglass, not only seemed to see canals on Mars but started seeing the things on a moon of Jupiter as well. Either there were a hell of a lot of alien civil engineers out there in the solar system (some with perhaps a long commute), or something was definitely wrong. Douglass suggested that “floaters” in the vitreous humor of the eye, or even structural factors in the outer surface of the eye itself, might generate false images. Douglass wrote to psychologist Joseph Jastrow for help with the psychological principles involved (apparently Jastrow never wrote back). But Douglass persisted. With W.H. Pickering, Douglass set up a disk, the same relative size through the telescope as the face of Mars, to try to make sense of the canal phenomenon (Hoyt 1976). Douglass studied “artificial planets” at various ranges to make sense of what he clearly suspected were optical illusions, and he later wrote an interesting paper (Douglass 1907) in which he catalogued a series of physical and psychophysical phenomena, many observed in his own experiments, in which he sought to explain the observation of what he referred to as “these faintest of markings,” the canals. Even in observing a blank white disk, Douglass repeatedly saw lines “radiating” from some point near the center of the given image, lines similar to those that also predominate Lowell’s 1896 drawings of the planet Venus, which seemed at the time to be infested with canali as well.

Not surprisingly, Lowell was none too crazy about Douglass’s essentially psychological explanation. Lowell believed the canals were physically real and reflected favorably on the grit and gumption of hard-pressed martian engineers, battling against the encroaching deserts of their ancient world (e.g., Amundson 2017). Douglass was essentially fired and went on to the University of Arizona, where he contributed enormously to the science of dendrochronology. Life and Lowell’s canals moved on. 

But that business with Douglass and the blank white disk should probably not be allowed to move on so quickly. The reason is that a blank white disk offers no Gestalt “anchors” on which to fasten canals by means of the usual suggested mechanisms, the Gestalt laws of closure and good continuation (King and Wertheimer 2005; Kohler 1947; Sharps 1993; Sharps 2018; Sheehan 1988). The only sources of “canali,” or of any other false perceptions or interpretations on a blank white disk, for that matter, would lie in the realms of individual differences and sociocognitive factors.

This led us to an interesting question: Could we make modern people see canals, or something equally false, with reference to these two types of influence alone, and on a blank white disk?

This was the subject of the following experiment. 

Methods and Details

The full moon at “supermoon” approach to Earth (close orbital approach) was photographed with a handheld Sony Alpha digital camera using a 70mm telephoto (“portrait”) setting. This resulted in a round, natural picture of an illuminated object in the night sky but with all surface detail blurred and over-illuminated to the point of invisibility; in short, we had a blank white disk, but it came from a natural source.

Using PowerPoint, we presented this image to forty-eight female and fifteen male college students (aged eighteen to twenty-two years; gender ratios reflected the classes available). These respondents were informed that they would view one of two things: either a “lunar or planetary object photographed by telescopic instruments within our solar system” (which was true) or a lunar or planetary object on which “authorities” in science and engineering had discerned features or structures, some of which may have been seen as “artificial” (which was also still true—there are all sorts of natural features on the moon, and you can always find the occasional PhD on cable TV who will yammer away on “documentaries” about buildings or even statues on the Moon).

We also administered the Dissociative Experiences Scale (DES), a measure of a critical individual-differences factor in peoples’ belief in the paranormal. In previous research published in SI (e.g., Sharps et al. 2016), we showed that subclinical dissociative tendencies as measured by the DES (tendencies to perceive the world as vague and potentially supernatural) not only biased ordinary people to believe in such things as aliens and Bigfoot but actually allowed them to see aliens and Bigfoot in stimulus pictures that were pretty obviously Halloween costumes to everybody else.

It is important to note that we are not talking here about dissociation in the clinical sense, the highest levels of which may be seen in dissociative identity disorder. In this experiment, we dealt with dissociation in the normal, subclinical, nonpathological sense experienced by everybody from time to time, in which the world seems slightly odd, but we don’t know why. If dissociation can contribute to false perceptions of Bigfoot, we thought, then it could certainly contribute to the observation of nonexistent features on other worlds. Dissociation, therefore, seemed the perfect individual difference to test in this context.

Results: Illusory Features of an Imaginary World

My research group has been privileged to publish articles in SI in which supposedly rational adults have demonstrated firm beliefs in Bigfoot, aliens, ghosts, and the 2012 return of the ancient Mexican god Quetzalcoatl. But we’ve never seen anything like this.

In the present study, 30.16 percent of our supposedly rational respondents provided completely imaginary accounts of structures they interpreted for us based on a blank white disk. These interpretations ranged from “fire which creates light” to “yellow and purple rays of light” to “many dots, evenly spaced” to “yellow object, mountain ranges, canal, hills, buildings” from a person who apparently wasn’t kidding. The interpretations were flying high and free. These were interpretations of a blank, white, round stimulus object. It could have been an illuminated golf ball. But about one-third of our respondents created false interpretations of what they believed they saw anyway.

Over one-third (34.92 percent) of our respondents reported seeing some kind of “structure” on this structure-free orb, and 25.40 percent reported colors, ranging from violet to yellow, on this purely white object. The color factor has precedent; Douglass (1907) reported many chromatic anomalies with his refractive telescopes. However, no such factors could have influenced these results. This was a simple white disk, presented by PowerPoint, with no chromatic aberrations inherent in the system. (No respondents identified this disk as the moon, although one person wondered whether it might be.)

The crux of these results: We presented reasonably educated people with a round white blob. About a third of them saw features, structures, and colors on this blob that weren’t there at all and provided wholly erroneous interpretations, essentially stories, about what they erroneously believed they had seen.

Results: Individual Differences in Dissociation

Did dissociative tendencies contribute to these bizarre perceptual and cognitive anomalies?

Absolutely. Regression analysis of DES scores versus number of interpretations yielded significance at p = .019, β = .442; the more dissociation, the higher the number of erroneous interpretations. Interestingly, this effect occurred only with the “planetary or lunar object” condition, not the “authorities” condition. Also, this effect occurred only with interpretation; dissociation did not contribute significantly to color or structural misperception.

This makes psychological sense. Dissociative tendencies would tend to operate at deeper levels of cognition. Color and structure perception are relatively “shallow” at cognitive levels; the work of Douglass, for example, is filled with references to color misperceptions that could as easily have had their origins in the chromatic aberration of lenses as in the psychology of the given observer (Douglass 1907). The same relatively shallow factor applies to structural shapes: either you see a round thing or a square thing—or you don’t.

But to decide that the given square thing is an alien headquarters or that the round thing is a flying saucer—that takes interpretation, the level at which we believe dissociative tendencies have their deeper, strongest, and most insidious effects. Dissociation not only convinces you that you believe in Bigfoot but that you see him—and this is exactly the misinterpretive effect of dissociation we see in the present data and that we have observed in previous studies (e.g., Sharps et al. 2014; Sharps et al. 2016).

Results: Sociocognitive Factors

There were no differences in the production of false colors, structures, or interpretations of our stimulus object between the “planetary/lunar” and “authorities” conditions at all. Also, dissociative factors only operated in the former condition. Why?

A useful theoretical perspective lies in Gestalt/Feature-Intensive Processing theory (G/FI; Sharps 2003; Sharps 2017; Sharps et al. 2014; Sharps et. al 2016). G/FI theory suggests a continuum between feature-intensive processing, which involves detailed analysis of specific features of the given situation, and Gestalt processing, in which those features are ignored in favor of faster, but less incisive, processing of wholistic concepts at the expense of detail-oriented analysis (also see Lakoff 1987).

In our “authorities” condition, scientists and engineers had reportedly already pronounced authoritative conclusions concerning alien presence and structures on the white, featureless disk; sociocognitively based Gestalt acceptance of these “expert” conclusions would therefore require relatively little feature-intensive analysis on the part of respondents. So we would expect to see little or no effect of dissociation on interpretation in this condition. People simply didn’t do the relevant feature-intensive interpretive processing.

However, in the lunar/planetary condition, our respondents were left without expert opinion to be accepted in a Gestalt manner—they were required to interpret a “lunar or planetary” stimulus on their own. Relying on their own intellectual resources for interpretation, those with dissociative tendencies tended to provide more and more outlandish, feature-intensive interpretations; they “saw” more evidence of the alien presence.

But regardless of dissociative tendencies, wouldn’t we expect respondents to generate more false interpretations with the sociocognitive backing of “authorities” in science and engineering?

This finding simply did not occur, and its absence brings up somewhat frightening questions: Is it possible that our respondents didn’t need authorities to support the false finding of alien influences on other worlds? Do modern people have an effectively Gestalt response to “lunar or planetary objects” that predisposes them to think in terms of science fiction and fantasy rather than science, of Grey Aliens and Klingons rather than orbital dynamics?

If so, we would expect that the removal of the “lunar or planetary” factor would dramatically decrease the number of false color and structural observations. We exposed an additional thirty-five college student respondents to a condition in which we simply told them they would see “an object,” removing all “lunar/planetary” and “authorities” references from the stimulus situation.

Removal of these astronomical references resulted in a dramatic decrease in interpretations, F (2.95) = 3.96, p = .022. Student-Newman-Keuls analysis (p<.05) revealed this was entirely the result of the addition of the “object” condition, which resulted in approximately 5 percent as many interpretations as did the “authorities” or “lunar or planetary object” conditions. These latter two conditions still did not differ significantly.

This result carries with it a sobering fact. Seeing our round white blob as an “object” yielded few false perceptions; however, seeing it as a “planetary or lunar object” yielded the same results as if “authorities” saw the same stimulus object as full of features and structures, or perhaps full of Klingons or little green civil engineers. 


As with our other studies cited here, the demographics of the student population for the present research rendered it reasonably representative of the current American population. These were essentially average people, and it would appear that average people, when confronted with astronomical phenomena, think much more in terms of Star Trek than of Neil DeGrasse Tyson. As soon as “planetary or lunar” concepts intruded on the given cognitive situation, people were much more likely to think in terms of aliens than of orbital dynamics, of spaceships and star bases than of physics and chemistry. Whether or not “authorities” had seen evidence of aliens on a given object in space, people were perfectly likely to generate the relevant science fiction/fantasy concepts, on a Gestalt basis, on their own.

The implications here are obvious and disturbing to those of us who prefer scientific and skeptical thought. These results indicate that people confronted with a lunar or planetary object are much more likely to associate it with science fiction than with science fact, without recourse to the admittedly complex realities of astronomy. Confronted with problems of planetary scope, most people don’t think about physics; they think about Darth Vader.

These results are consistent with the hypothesis driving this research, the idea that cognitive processes, configured by sociocognitive and dissociative factors, determine to a great degree the average person’s perception and understanding of scientific and pseudoscientific phenomena. The canals of Mars did not exist—but we can make them in the minds of respondents to carefully crafted experimental stimuli.

These results very strongly support the importance of education of the type advocated by the Committee for Skeptical Inquiry. As a modern culture (and species), we are desperately in need of better education in the sciences, including those sciences that deal predominantly with planetary and astronomical realities. In the absence of such education, this research indicates that the average person is likely to start thinking about Vulcans or Captain Kirk whenever the word planet is mentioned. These findings bode poorly for an increasingly science-intensive future, unless we as a society can begin to turn more toward fact at the expense of fantasy.


  • Amundson, M.A. 2017. Seeing Arizona, imagining Mars: Deserts, canals, global climate change, and the American West. Journal of Arizona History 58: 331–350.
  • Brasch, K. 2018. Canal mania. Sky and Telescope 136: 28–33. 
  • Douglass, A.E. 1907. The illusions of vision and the canals of Mars. Popular Science Monthly 70: 464–474. 
  • Evans, J.E., and E.W. Maunder. 1903. Experiments as to the actuality of the “canals” observed on Mars. Monthly Notices of the Royal Astronomical Society 63: 488–99.  (Note: this article is frequently cited incorrectly with regard to the order of authors and the pagination.)
  • Hoyt, W.G. 1976. Lowell and Mars. Tucson: University of Arizona Press.
  • King, D.B., and M. Wertheimer. 2005. Max Wertheimer and Gestalt Theory. New Brunswick: Transaction.
  • Kohler, W. 1947. Gestalt Psychology. New York: Mentor.
  • Lakoff, G. 1987. Women, Fire, and Dangerous Things. Chicago: University of Chicago Press. 
  • Sharps, M.J. 1993. Gestalt laws of perceptual and cognitive organization. In Magill’s Survey of Social Science: Psychology. Pasadena, CA: Salem Press.
  • ———. 2003. Aging, Representation, and Thought: Gestalt and Feature-Intensive Processing. Piscataway, NJ: Transaction Publishers.
  • ———. 2017, 2nd ed. Processing Under Pressure: Stress, Memory, and Decision-Making in Law Enforcement. Flushing, NY: Looseleaf Law.
  • ———. 2018.  Percival Lowell and the canals of Mars. Skeptical Inquirer 42(May/June): 41–46. 
  • Sharps, M.J., S.W. Liao, and M.R. Herrera. 2014. Remembrance of apocalypse past: The psychology of true believers when nothing happens. Skeptical Inquirer 38(November/December): 54–58.
  • ———. 2016. Dissociation and paranormal beliefs: Toward a taxonomy of belief in the unreal. Skeptical Inquirer 40(May/June): 40–44.
  • Sheehan, W. 1988. Planets and Perception. Tucson, AZ: University of Arizona Press.