A LIVELY ELECTRONIC COMPENDIUM OF RESEARCH, NEWS, RESOURCES, AND OPINION
Astronomy Education Review
2013, AER, 12(1), 010202, http://dx.doi.org/10.3847/AER2013007
Astronomy Exercises for the Artist: van Gogh the
Timothy M. Lawlor
Pennsylvania State University—Brandywine, Media, Pennsylvania 19063
Received: 03/25/13, Accepted: 06/4/13, Published: 06/24/13
C2013 The American Astronomical Society. All rights reserved.
We present a set of exercises designed to be used in a survey astronomy course, an introductory astronomy
laboratory course, or in secondary education. The exercises use the great works of Vincent van Gogh but could
be easily extended to other works of art. We also include a brief description of our current practices, lectures, and
group questions as examples of material that should be discussed prior to use of these “AstroArt” exercises.
1. INTRODUCTION AND LEARNING OBJECTIVES
One objective of this work is to help to provide additional motivation and extend interest and appreciation for the
study astronomy to non-science students. We aim not only to engage students in learning astronomical content
and concepts but also to engage them in the broader endeavor of asking questions, seeking appropriate and
veriﬁable evidence, revising questions when needed and in general learning how to learn regardless of the topic
We use four paintings of van Gogh to formulate questions that require knowledge of basic astronomy concepts. It
is generally believed that van Gogh was a keen observer and often painted accurate depictions of scenery
including on occasion the moon and the night sky (Whitney 1986;Dye 1985). This idea has indeed found real
use in solving historical mysteries in van Gogh’s White House at Night (Olson et al. 2001), Edward Munch’s
The Scream (Olson et al. 2007), and Ansel Adams’s Autumn Moon (Olson et al. 2005). Earlier, Whitney (1986)
and Boime (1984) carefully analyzed the astronomical signiﬁcance of van Gogh’s night sky paintings, including
those that we use here. Both of these references serve as excellent guides for comparison to the answers we
provide to our exercises. We include those comparisons in this work. Soth (1986) wrote a thorough analysis
focused mainly on van Gogh’s painting the Starry Night. This work provides some astronomical investigation,
though focuses more on van Gogh’s motivation and religious inﬂuences on the composition. We also refer to the
on-line van Gogh Letters Project (http://vangoghletters.org/vg/letters.html, referenced henceforth as “Letters
Project”). Our exercises based on van Gogh’s paintings are intended to explore the phases of the Moon including
time of day and direction of observations, constellation identiﬁcation, and the daily and yearly motion of the sky.
They require and promote active learning and research through the use of on-line resources and a planetarium
program which aid in answering questions, solving problems, and conﬁrming or reexamining answers. They can
be done in a lecture setting without online access, but information found on-line would have to be provided as
given clues for student groups. Some questions do require computer access for identifying constellations for
example, thus may be better suited for a laboratory exercise. The reasoning used in these exercises may negate a
hypothesis or expose ﬂaws in astronomical artistic depictions as well.
We leave the choice of teaching technique open to the reader. The author has generally used an adaptation of the
well-known “Think, Pair, Share” teaching technique (see for example; Slater et al. 2003;Forestell et al. 2008;
Prather and Brissenden 2008) or Problem Based Learning (Shipman and Duch 2001; http://www.udel.edu/inst/).
They can also be adapted for younger learners using, for example, “Claims, Evidence, Reasoning” as described
by McNeil and Kragcik (2012).
In Section 2, we review the groundwork of topics and concepts that should have already been discussed before
attempting these “AstroArt” exercises. In Section 3, we present the exercises themselves including sample
questions and answers, and we compare our answers to previous published studies of paintings. We intend to
maintain a database of astronomical facts for any painting and we discuss this brieﬂy in Section 4. Finally, we
provide a brief summary and discussion in Section 5.
2. THE GROUNDWORK
In this section, we brieﬂy describe examples of mini-lectures and group questions that should be completed
before the Astroart exercises are attempted. The broad topics covered here are the yearly motion of the sky, the
motion of the Sun, the phases of the Moon, and the stars and constellations. The yearly motion of the sky is
introduced in a usual way, with a diagram of the Earth and Sun from outside the system. Particularly, we are
careful to distinguish the night and day sides of the Earth, and couple this with concepts relating to Earth’s orbit,
rotation, the ecliptic, the zodiac and circumpolar constellations. Once students understand why we see different
constellations during different seasons, we break into groups to discuss why it is that Ursa Major, for example,
can be found in the night sky all year long from our geographic location in the northern hemisphere. These
preliminary exercises are useful for a variety of reasons. For one, they introduce the concept of observation
relative to location on Earth and relative to time of day on Earth. This leads naturally into discussion of reason
for the seasons and a similar model of the system is used for discussing the phases of the Moon.
Here, we present two sample questions that pairs or small groups of students discuss during lecture class,
typically using an impromptu version of “Think, Pair, Share.” For example, the instructor lists three or four
different answers on the board, and students are asked to vote by holding one, two, three or four ﬁngers up in
below their chin to vote on their answer. This allows the instructor to decide in real time if the topic needs more
time for discussion and makes it difﬁcult for students to see each other’s answers, eliminating the discomfort of
giving the wrong answer. Some sample questions related to the yearly changing night sky are as follows:
A close friend asks you to proof read a poem that he wrote for a creative writing class. The poem meanders on
and on about a walk along the beach and makes note of how the constellation Orion seems to follow him along
the coast no matter what time of year it is. You are forced to correct your friend— how and why?
Unfortunately in the same poem, your friend describes in great detail watching the Little Dipper set over the
ocean down on the Jersey Shore. What are the problems here?
The yearly motion of the sky also gives rise to the reason for the seasons and why and where we can experience
24 h of light or darkness. An example of diagrams used in mini-lectures to discuss this topic is shown in Figure 1.
Before beginning to study the reason for the phases of the Moon, we lead with group questions based on Figure 2,
which is designed to further cement the concept of observational perspective. It cannot be overstated how
Figure 1. Sample diagrams for discussing the reason for the seasons, and other implications. Obviously this is not drawn to
scale. The size of Earth is exaggerated to make it more clear what causes the seasons and to point out the tropics and circles.
This model introduces the terminator between night and day, which introduces an important concept in discussing the phases
of moon later
difﬁcult this tends to be for students to grasp. Even given these simple diagrams, it is often difﬁcult for some
students to predict what the ball will look like from the observer’s position. This uncovers a fundamental issue in
trying to teach the reason we see phases of the Moon, which is a more complicated situation than this example. In
the ideal case of being in a laboratory course, or having a small lecture class (which is probably uncommon), this
exercise can be carried out with a bright lamp or overhead projector and a ball. We have found that racquetballs
show the terminator better than other choices.
Finally, in Figure 3, we show an example of the model that we use for discussion the phases of the Moon. Note
that in this model, we remove the tilt of the Earth relative to the ecliptic, ignore the inclination of the Moon’s
orbit, and approximate the orbital period of the Moon. From this diagram, one can extract three pieces of
information: Lunar phase, approximate direction of observation, and approximate time of day of observation.
Given any two of these, the third can be uniquely determined. The corresponding mini-lecture that goes along
with this slide consists of eight of these diagrams, with each phase being discussed one by one. Again, this is
only a brief summary. We use many other resources, such as applets, pictures, and descriptions, when
approaching this difﬁcult to teach topic.
Next, we present three lunar phase related questions used for group discussion:
Figure 2. Putting observing into perspective and perspective into observing
Figure 3. Sample model for discussing the phases of the Moon. An interesting discussion point is to explain why the portion
of the visible moon points away from the Sun for the waning phase pictures. Two other difﬁculties arise from this model. One
is that “south,” as deﬁned for an observer in the northern hemisphere, is not the same as for our observer shown on the equa-
tor. A moon high in the southern sky for a northern observer would instead be best described as directly overhead by an ob-
server on the equator. Second, the 2-D depiction sometimes creates confusion in understanding what one can or cannot see
when looking toward north from the position on the equator. For example, when asked what direction one should look to see
the last quarter moon at sunset (which is a trick question) students occasionally answer “north,” as if the Earth is a ﬂat disk
that can be looked through
Out for a walk one evening, you have become a bit turned around. You were supposed to head west on Broad
Street and cannot recall which way is west! It is the early evening and the Sun has already set. Scrambling for a
clue, you look up in the sky and see a thin crescent moon close to the horizon. How can you use all of this
information to walk in the right direction?
If your friend tells you that they have seen a full moon in the western sky, what time of day can you conclude that
they were viewing the Moon?
In what direction will you see the Moon if it is in the waxing crescent phase at sunset?
By this time during the semester, students would also have been exposed to basic constellation recognition and if
in a lab course, they would have extensive hands on use with a planetarium program such as STARRY NIGHT
BACKYARD or STELLARIUM. Lecture students will be at least visually familiar with this software. Such software is an
important tool for some of the questions presented in the following section. Topics explored during lecture using
planetarium software include the daily motion of the sky, circumpolar constellations, zodiac constellations,
inaccurate astrological signs, and yearly motion of the night and daytime skies. We note that STELLARIUM is an
open source software (i.e., free to download).
3. ASTROART EXERCISES WITH SAMPLE QUESTIONS AND ANSWERS
A number of van Gogh’s paintings depict scenes that include observations of the night sky. During the spring and
summer of 1888, the artist became preoccupied with painting a starry sky. In a letter dated on April 19 to Emile
Bernard (Letters Project), van Gogh wrote “A starry sky for instance—look that is something I should like to try
to do.” On June 18 of the same year, he writes again to Bernard: “But when shall I paint my starry sky, that
picture which preoccupies me continuously?” He did get around to painting two starry night scenes in September
1888 and the iconic Starry Night in June 1889. He included the Moon in other various paintings as well. In some
of van Gogh’s night sky paintings, stars and objects are immediately recognizable, but we can ﬁnd out more form
his observations. Though we only include van Gogh in this set of exercises, there are likely many more artists to
be used across a variety of mediums. Here, we present four of van Gogh’s well known paintings and of course
we begin with the Starry Night.
3.1. The Starry Night—St. Remy-de-Provencal, France, June 1889
What is the phase of the Moon that Vincent van Gogh depicts in this painting?
The phase shown is a waning crescent phase (more on this below) (Figure 4).
Figure 4. The Starry Night
What time of day would you predict van Gogh observed this scene? What evidence can you cite to support your
The time of day depicted is early morning. The primary evidence is that the left side of the moon is illuminated
and the moon is close to the horizon. Another way to determine this is by using the Earth-Moon-Sun diagram
shown in Figure 3.
Approximately what direction was van Gogh facing when he painted this scene?
Van Gogh was looking toward the east. Because the waning moon is rising, it has to be in the east direction (also
see Figure 3). The illuminated part of the crescent moon always “points” toward the Sun. In this case it is the
rising Sun, which rises in the east.
If a planetarium program is available, and after searching online to ﬁnd approximately when and where van
Gogh painted this scene, determine as best you can the date and range of time of day that van Gogh may have
observed this scene. Do not forget that you must set the correct geographic location in the software. Record all
references for collected data. Using the planetarium software, what constellations and other objects can you
identify in this painting?
From a rudimentary on-line search, one can ﬁnd that van Gogh observed this scene in June 1889 in Saint-R
de-Provence, France. Deciding an exact time and date and identifying stars is problematic because we cannot tell
the exact scale that van Gogh used. In fact, the Alpilles foothills shown in the painting are known to be
exaggerated in steepness of slope (Soth 1986). But based on the time of day and the location of the waning
crescent moon relative to the horizon, it is possible to identify Venus to the right of the foreground Cypress tree.
Whitney (1986) also identiﬁes the phase as crescent in the easterly direction and suggests from van Gogh’s
letters that Venus was observed in May. If we estimate by the orientation of the Moon and Venus, we can ﬁnd a
range of time of day using a planetarium program. Based only on this celestial orientation, a good guess is on
June 21, 1889 between 3:45 am and 4:10 am. We note here, however, that although this is visually a very good
match, it has been proven (Whitney 1986;Soth 1986 and references therein) that van Gogh completed the Starry
Night before on June 18, 1889. This is based on van Gogh’s own words referencing the painting in a letter to his
brother Theo (Letters Project) dated on or about June 18: “At last I have a landscape with olive trees, and also a
new study of a starry sky.” Both Whitney and Soth conﬁdently report that the painting was completed between
the dates on June 16–18 based on van Gogh’s letters. Speciﬁcally, they cite that he did not mention it among
completed works in a letter to his sister dated on June 16 and then did refer to it in a letter to Theo on June 18
(Letters Project) as quote above. Although this is strong evidence for the exact date being between 16 and 18, it
may be possible if not likely that van Gogh simply did not mention it in the letter to his sister. Perhaps, he saved
the description of a starry night for his brother? In any case, this date range is generally accepted to be the range
for the completion of the painting. Boime (1984) identiﬁes the date as on June 19. Even so, on all of these dates
the moon was in the waxing gibbous phase, which is clearly not what is depicted in the Starry Night. Boime
suggests that perhaps van Gogh changed his mind and artiﬁcially changed the gibbous moon to crescent, though
there is no real compelling evidence for this.
If we take a leap, and assume that van Gogh painted this scene in June but observed it at the end of May, which
is at least implied by Whitney, one may guess that the scene was observed on May 25 around 4:30 am. On that
morning, Venus and Aries are a bit closer, but again there is no concrete evidence for this claim. This date does
show the complete sky though, including the crescent moon. It may be possible that Soth offers the right
interpretation for the discrepancy in phase. Soth writes that van Gogh painted the crescent moon, church and
village from a memory of the painting “Fesival of St. John,” by Jules Breton. Soth believes that Starry Night in
an ‘imaginative amalgam’ from both memory and local scenery. In any case, for the sake of this exercise, the
arrangement of the Moon and Venus may lead students to conclude that on June 21 (or so) is the “right” date,
even though it was certainly painted before June 18.
It is difﬁcult in this painting to identify exact stars or constellations, but one could well argue that the stars above
Venus and the tree belong to the constellation Aries, while the stars just to the left of the Moon may belong to the
constellation Pisces. The identiﬁcation of Aries is in agreement with the work of Boime (1984). In contrast to
Boime, Whitney does not believe that the constellations can be identiﬁed and goes on to suggest that an entirely
different part of the sky, including the glow of the Milky Way and the constellation Cygnus, have been
transposed onto the scene. Given Soth’s convincing discussion about van Gogh’s shift to using imagination to
depict nature and make it better than reality, it is possible he moved the sky.
3.2. Road with Cypress and Star—April–May 1890
What phase of the Moon is shown in this painting?
The phase of the Moon shown is a waxing crescent. The right side of the moon is illuminated indicating that it is
waxing. This is conﬁrmed by Whitney (1986) who reports that the scene was two to four evenings after new
moon. Whitney also notes that the tilt of the moon is slightly off in that it is parallel with the horizon which
would indicate the Sun had not yet set (Figure 5). Van Gogh himself offers an interesting topic for discussion
about a common misconception about the reason for the phases of the moon. In a letter dated on June 17, 1890 to
Paul Gauguin (Letters Project), van Gogh writes:
“A last try—a night sky with a moon without brightness, the slender crescent barely emerging from the opaque
projected shadow of the earth—a star with exaggerated brightness, if you like, a soft brightness of pink and green
in the ultramarine sky where clouds run.”
Seemingly van Gogh is suggesting that the shape of the crescent moon is caused by the shadow of Earth, thus
confusing it with a lunar eclipse. This very common misconception persists today.
Approximately what time of day does this scene depict and what evidence do you have for your claim?
This scene is during the evening, shortly after sunset. The telltale signs of this are that the Moon is in the waxing
crescent phase and is fairly close to the horizon. These two observations allow one to use Figure 3to predict the
time of day. Also, the illuminated portion of the crescent moon is “pointing” toward the setting Sun, thus it is
In what direction was van Gogh facing to paint this scene?
Because the illuminated portion of the crescent moon “points” toward the setting sun, we can conclude that van
Gogh was (roughly) facing west. This is again supported by Whitney (1986) who determines van Gogh was
Figure 5. Road with Cypress and Star
If a planetarium program is available, and after searching online to ﬁnd approximately when van Gogh painted
this scene, determine as best you can the date and range of time that van Gogh may have observed this scene.
Can you identify the bright object to the left of the cypress tree? Do not forget that you must set the correct
geographic location in the software.
We can ﬁnd from on-line resources that this painting was done during May 1890, about three months before the
artist’s untimely death (see, for example: http://www.vangoghgallery.com/misc/chrono.html). It can also be
found that this was done in the evening, which is clear from the lunar phase, but also from letter RM23 in the van
Gogh Letter Project (http://vangoghletters.org/vg/letters/RM23/letter.html). Here again scale is tricky, and thus
the exact time in the evening is also approximate. A possible identiﬁcation for the star, time of day, and date
include: April 22–23 between 8:00 pm and 8:45pm: If this is the case the star to the left of the waxing crescent
moon could be the armpit of Orion the Hunter, Betelgeuse. This is consistent with the analysis of Whitney
(1986), who makes this same suggestion. The star could be Sirius if the scale is greatly exaggerated and
Betelgeuse is behind the foreground tree. Whitney (1986) concludes that the painting was completed between
May 13 and 16 but observed between April 21 and 23. This conclusion is made because April 19 would have
been the last New Moon that occurred, while van Gogh was still in St. Remy, before he left. Whitney goes
further to settle on April 21 based on meteorological records, which would make the crescent moon only two
days old. The orientation of the Moon and star (assuming we settle on Betelgeuse) on April 21 is not as close to
the orientation in the painting as is on April 22, but it is hard to argue with meteorological records. The
orientation on the 21 is fairly good for the star Rigel which is also the quite bright, and the bright star Aldebaran
is also nearby. The orientation for Aldebaran is also not quite right.
Based on the “right” orientation only, a student may choose on May 22 between 8:15 pm and 9:00 pm: In this
case, they may predict that the star is Procyon, whose apparent magnitude is approximately þ0.04. One might
argue that it is a bit later than 9:00 pm because earlier than that, Venus would be quite visible. However, this date
in reality is ruled out (see for example, Hulsker 1990). Further evidence for ruling it out arises from letter 873
from van Gogh to his brother Theo (Letter Project) dated on May 20. Van Gogh wrote this letter from Auvers-
sur-Oise. Thus, van Gogh had already been discharged from the hospital of St. Paul-de-Mauole and had left St.
Remy where this painting was composed before on May 22.
Finally, it may be suggested that the “stars” in this scene are Venus and Mercury which appeared in the sky on
April 21, 1890 until at least 8 pm, but for this to be the case both planets would have had to have been switched
to a mirror image, vertically about the Moon. Both appeared to the right, not the left, of the moon on that
evening. This would be an interesting claim for a student to make, however, Mercury is usually quite dim and
this was just after sunset, making it a dubious claim in the opinion of the author.
e Terrace at Night—Arles, France, September 1888
If a planetarium program is available, and after searching online to ﬁnd approximately when van Gogh painted
this scene, can you identify the stars or constellation in this painting?
In what direction was van Gogh facing? Do not forget that you must set the correct location in the software.
Record all references for collected data. During what time of year was this scene painted? What evidence do you
have to support your claim?
One can ﬁnd that this painting was painted in Arles, France (letter 678, Letter Project, for example). In fact, the
e still exists and has been refurbished to look the same as it did in van Gogh’s painting. Given this landmark’s
known location, we can ﬁnd that van Gogh faced south. One way to do this is to use Google Maps street view to
virtually visit the caf
e. Unfortunately, you can only view it from about a block away, as Google does not image
this speciﬁc street. But between the view from a block away and current day pictures it is possible to tell that the
e is on the east side of the street. The direction is also conﬁrmed in Boime (1984). From van Gogh’s
perspective, on the left or east side of the street, his view appears to be slightly southwest (Figure 6).
As a disclaimer, the author admits that he has always thought that the constellation depicted here looks like the
constellation Scorpio, with Antares being the star on the top left and the stars on the right completing Scorpio’s
familiar “V” shape. The stinger would extend out of site behind the buildings on the left. In a letter to his
sister, van Gogh describes how he was interrupted in writing his letter to her by painting this scene. That
letter was written beginning on September 9 and continued on September 16, 1888. During this time in the
evening, the constellation Scorpio does appear just west of south. Interestingly, at that same time, the very
bright planet Jupiter was situated just to the right of the constellation. And further, Mars moved directly
through this constellation during those dates. For Scorpio to be the right guess, the observation of the sky
would have had to have happened during civil twilight, around 7:15 pm. It could be that the sky was
observed or painted earliest before the street scene, and the sky view would have of course changed by the
time the painting was complete. If this is the case, we can rule out on September 11, since the moon would
have been dominating the constellation on that evening. Although the claim is unfounded, it is worth noting
that the space between the buildings is rather narrow and so the stars would have almost certainly changed
over the course of the completion of the painting, so the observation of the stars and sky may not be from
the same time as the caf
e in the foreground. If the two brighter ‘stars’ on the right side of the painted sky
are Jupiter and Mars, it could explain why these are a bit brighter than the quite bright star Antares. In fact
between September 11 and 13, Jupiter and Mars were very nearly on top of the right side of the “V” of the
constellation. If this is not the deﬁnitive answer, it is worth considering that these bright objects may have
been the inspiration for the scene. Considering this answer, some possible dates and times for the sky are
September 10 between 7:15 pm and 8:30 pm, or slightly later on September 12–16. The later dates allow for a
darker scene, but Scorpio moves further southwest making it less likely, unless the sky was observed earlier. In
this composition, Whitney (1986) does not believe that the stars can be convincingly identiﬁed and that they
have no real basis in the sky. On the other hand, Boime (1984) seems convinced that the scene was observed
around 11 pm, based on the lack of caf
e customers, and that the stars are part of the constellation Aquarius. The
author ﬁnds this an unlikely answer primarily because the brightest stars in Aquarius have an apparent magnitude
between about 3.0 and 4.0 and so are not particularly bright. This would especially be a problem close to the caf
Another possibility is that the scene was later in the evening (8:00 pm to 9:00 pm) and these stars belong to
the “tea pot” stars in the constellation Sagittarius. This would rule out on September 13 and 14 because the
moon would have dominated that part of the sky and would have been quite bright, in a later phase. The
positions of the stars are similar (with some imagination) as the spout and top of the lid of the teapot,
thought the relative brightness of the stars is wrong. Here again, none of the stars are all that bright, with
magnitudes between about 1.8 and 3.0. The brightest star in the painting would be the one with the highest
magnitude (or the least bright).
Figure 6. Caf
e Terrace at Night
3.4. Starry Night over the Rhone—Arles, France, September 1888
Can you identify which constellation is shown in this night sky scene?
Immediately identiﬁable, this is the constellation Ursa Major. It contains the most easily identiﬁable asterism,
The Big Dipper.
What direction would you predict that van Gogh was facing while viewing this scene? What evidence can you
cite to support your claim?
Astronomically, one may predict that van Gogh was facing north. The evidence for this is that the front two stars
of the dipper are used as pointers to the North Star, Polaris. In this case, they point straight up. Note here that it
has been shown in numerous places that van Gogh was actually facing southwest and transposed the northern
constellation onto the sky artiﬁcially. One can witness this fact directly using Google Maps street view to view
the relatively recognizable scene, including the Trinquetaille bridge, by virtually standing on the street Chemin
des Segomaux just south of Place Lamartine in Arles, France. On this street, one must face southwest in order to
see the scene (Figure 7).
Although not speciﬁcally included in these exercises, Whitney (1986) performs an excellent exercise that would
be well suited for a lecture or lab at this level. By using the Big Dipper, he approximates the scale of angle on the
sky. Using his approximation, he estimates that the bowl of the Big Dipper is about 11from the horizon, and
knowing that the North Star is 33above the bowl he ﬁnds that Polaris is about 44above the horizon—an angle
very nearly the same as the latitude of Arles, France. This demonstrates the usefulness of estimating angles in the
sky, a typical Astro 001 topic, as well as it demonstrates that Polaris is an angular height above the horizon that
is roughly equal to an observer’s latitude. It is also a testament to the carefulness with which van Gogh painted
the stars over the Rhone.
If a planetarium program is available, and after searching online to ﬁnd approximately on what date van Gogh
painted this scene, can you conﬁrm your choice of constellation and direction? If so, approximately what time of
day did van Gogh observe this constellation? Describe all evidence and reasoning that lead you to your
Ignoring the true direction of the scene, we can still guess at what time of day van Gogh would have had
observed the constellation by observing its orientation in the sky (upright and on the horizon). Van Gogh really
did view the Big Dipper, as is indicated in letter 691 (Letters Project), in which he refers to Ursa Major, the
“Against the green-blue ﬁeld of the sky the Great Bear has a green and pink sparkle whose discreet paleness
contrasts with the harsh gold of the gaslight.”
Figure 7. Starry Night over the Rhone
Here, his description of a “green and pink sparkle” is likely due to atmospheric diffraction and conﬁrm the stars’
location low on the horizon.
One can ﬁnd that this painting was also done in Arles France during the same month as the Caf
e Terrace at
Night, September. In the same letter to his sister, van Gogh mentions that “At present I absolutely want to paint a
starry night.” Since the Caf
e Terrace at Night was done between the 12 and 16, we conclude that this was
observed sometime after the 16. By primarily using the orientation of the Big Dipper in van Gogh’s sky, we
estimate two dates over a range:
On September 20, the constellation would have been viewed in this orientation between 10:30 pm and 11:15 pm.
If the observation were later, say on September 27, the time of night would have been a bit earlier between
10:00 pm and 10:45 pm. This is in line with the ﬁnding of Whitney (1986). Boime compares the painting to the
sky at 9:00 pm and suggests that this indicates that van Gogh tilted the constellation upward. It is not
immediately clear why he suggests this, but 9:00 pm does not compare well to the painting.
4. FUTURE DATABASE
One additional motivation for these exercises is to create and maintain a database for collected information about
well known, or even less known, works of art in any medium that includes any accurate astronomical scene. It
will also be possible to keep track of semi-accurate examples, as well as indicate those that are not real
depictions. In this regard, student work can be used to ﬁne tune the details of the observations. The database will
include items such as identiﬁable objects and constellations, time of day, possible dates, direction of observation,
and lunar phase if applicable.
5. SUMMARY AND DISCUSSION
We have presented a sample of astronomy exercises that take advantage of the careful and keen eye of Vincent
van Gogh. They encompass what would typically be the ﬁrst two or three chapters of a survey college level
astronomy course, using concepts such as phases of the moon and how the sky moves daily and annually. For the
non-scientist, these topics can be particularly challenging and many common misconceptions continue. For an
impressive list of common misconceptions, see as an example, http://www.physics.umaine.edu/ncomins/. When
Whitney and Boime undertook their studies of van Gogh’s skies, they utilized connections at a full scale
planetarium to recreate the sky scenes. Thanks to easily accessible planetarium software, one can retrace their
steps, think critically and draw their own conclusions. We hope that applying concepts of astronomy to the
artistic masterpieces of van Gogh will provide additional motivation and interest to non-science students. Indeed
the opposite is true that astronomy in art has raised the author’s interest in art. I was lucky enough recently to
ee d’Orsay in Paris and view Starry Night over the Rhone in person. In person, the composition and
colors are absolutely breathtaking. The author ﬁnds it remarkable that such deep interest in the philosophy,
motivations and thoughts of van Gogh have persisted for more than a century after his premature death. This
seems especially so for his paintings of the night sky. It may be that the cyclic nature of the sky provides a
familiar pallet and feels like a connection to the past. Certainly, van Gogh has sent his familiar pallets forward to
generations and for that the author is very grateful.
The author would like to thank an anonymous referee for pointing to important references and making
suggestions that have resulted in a much improved manuscript.
Boime, A. 1984, “Van Gogh’s Starry Night: A History of Matter and a Matter of History,” Arts Magazine,
Dye, L. 1985, “Starry-Eyed: But van Gogh’s Feet were on the Ground, Astronomers say,” in Los Angeles
Times, 17 January.
Forestell, A., Brissenden, G., Prather, E., and Slater, T. 2008, “Revisiting Think-Pair-Share: An Expanded
‘How-To’ Guide,” in This Month’s Teaching Strategy, March 2008, The Center for Astronomy Education web-
Hulsker, J. 1990, Vincent and Theo Van Gogh: A Dual Biography, First English language edition, Ann Arbor:
Fuller Technical Publications.
McNeil, K. L., and Kragcik, J. S. 2012, Supporting Grade 5–8 Students in Constructing Explanations in
Science, Boston: Pearson Education, Inc.
Olson, M. S., Olson, D. W., and Doescher, R. L. 2007, “The Blood-Red Sky of Munch’s The Scream,”
Environmental History, 12(1), 131.
Olson, D. W., Doescher, R. L., Burleson, J. A., Davidson, H. E., Denkeler, L. D., FitzSimon, E. D.,
McGillicuddy, R. P., Montondon, D. N., Sanchez, T., Voss, V. A., Walker, J. L., and Wells, A. E. 2001,
“Identifying the ’Star’ in a Long-Lost van Gogh,” Sky & Telescope, 101(4), 34.
Olson, D. W., Doescher, R. L., Holsinger, K. D., Ralph, A. B., and Valencia, L. D. 2005, “Ansel Adams and an
Autumn Moon,” Sky & Telescope, 110(4), 40.
Prather, E., and Brissenden, G. 2008, “Development and Application of a Situated Apprenticeship Approach to
Professional Development of Astronomy Instructors,” Astronomy Education Review, 7(2), http://aer.aas.org/
Shipman, H. L., and Duch, B. J. 2001, “Problem-based learning in large and very large classes,” in The power
of problem-based learning: A practical ‘how to’ for teaching courses in any discipline, ed. Duch, B. J., Groh, S.
E., and Allen, D. E., Sterling, VA: Stylus, 149.
Slater, T. F., Prather, E. E., Bailey, J. M., and Adams, J. P. 2003, “Effectiveness of a Lecture-Tutorial-approach
to large enrollment introductory astronomy,” in 76th International Conference of the National Association for
Research in Science Teaching, Philadelphia, PA, Abstract available in NARST Program and Abstracts—76th
NARST International Conference 2003, 226.
Soth, L. 1986, “Van Gogh’s Agony,” The Art Bulletin, 68, 301.
Whitney, C. A. 1986, “The Skies of Vincent van Gogh,” Art History, 9(3), 351.