Astron. Nachr. / AN 999, No. 88, 789 – 795 (2011) / DOI please set DOI!

An Arabic report about supernova SN 1006 by Ibn Sı̄nā (Avicenna)

R. Neuhäuser1 ? , C. Ehrig-Eggert2 , and P. Kunitzsch3
1
Astrophysikalisches Institut und Universitäts-Sternwarte, FSU Jena, Schillergäßchen 2-3, 07745 Jena, Germany (e-mail:
rne@astro.uni-jena.de)
2
Institut für Geschichte der Arabisch-Islamischen Wissenschaften, Westendstraße 89, 60325 Frankfurt a.M., Germany
(retired)
3
LMU Munich, Germany (retired); home: Davidstrasse 17, 81927 München, Germany
arXiv:1604.03798v2 [astro-ph.SR] 3 May 2016

Received 2016 Feb 17, accepted 2016 Apr 8

Published online

Key words supernova - SN 1006

We present here an Arabic report about supernova 1006 (SN 1006) written by the famous Persian scholar Ibn Sı̄nā (Lat.
Avicenna, AD 980-1037), which was not discussed in astronomical literature before. The short observational report about
a new star is part of Ibn Sı̄nā’s book called al-Shifā’, a work about philosophy including physics, astronomy, and meteo-
rology. We present the Arabic text and our English translation. After a detailed discussion of the dating of the observation,
we show that the text specifies that the transient celestial object was stationary and/or tail-less (a star among the stars),
that it remained for close to three months getting fainter and fainter until it disappeared, that it threw out sparks, i.e. it
was scintillating and very bright, and that the colour changed with time. The information content is consistent with the
other Arabic and non-Arabic reports about SN 1006. Hence, it is quite clear that Ibn Sı̄nā refers to SN 1006 in his report,
given as an example for transient celestial objects in a discussion of Aristotle’s Meteorology. Given the wording and the
description, e.g. for the colour evolution, this report is independent from other reports known so far.

c 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1 Introduction: Supernova 1006 brightness would then be −7.5 ± 0.4 mag (Winkler et al.
2003).
SN 1006 was observed by the Yemeni observer(s)
Historic observations of supernovae (SN) are important to
around Apr 17/18 (Rada & Neuhäuser 2015), by c Alı̄ ibn
understand SNe, neutron stars, and SN remnants (SNR):
Rid.wān since Apr 30 (Goldstein 1965; SG02), and in China
Historic reports can deliver the date of the observation
and Japan since the end of April or early May (SG02). The
(hence, the age of the SNR and, if existing, of the neutron
positional information by c Alı̄ ibn Rid.wān (ecliptic longi-
star) together with a light curve (hence, possibly the SN
tude range) led to the identification of the SNR (together
type), sometimes the colour and its evolution, and the po-
with the right ascension range from the Chinese and the dec-
sition of the SN, which is needed to identify the SNR and,
lination limit from St. Gallen), see Stephenson et al. (1977)
if existing, the neutron star and/or pulsar wind nebula. Such
and SG02. Several Arabic observers noted stationarity. The
historic observations have been used very successfully for
report of Ibn Abı̄ Zarc (died in or after AD 1326) from a Mo-
SNe 1006 (from Eastern Asia, Arabia, and Europe), 1054
roccon source about SN 1006 (Goldstein 1965) – based on
(from Eastern Asia and Arabia), 1181 (only from Eastern
the edition of the Arabic and Latin text by Tornberg (1843)
Asia), and SNe 1572 and 1604 (from Eastern Asia and Eu-
– is the only source possibly mentioning a day-time obser-
rope), plus a few more SNe from the 1st millennium AD
vation: Its appearance was before sunset ... (SG02).
(see Stephenson & Green 2002, henceforth SG02, and ref-
erences therein). While the Arabic report about SN 1054 The following Arabic terms were used for historic ob-
merely confirms a bright new star in Gemini/Taurus around servations of SNe:
AD 1054, the Arabic reports about SN 1006 present a lot of – kawkab, which means star or planet, or more generally
detailed information (Goldstein 1965; Cook 1999; SG02; celestial object, used e.g. for SN 1006 by Ibn al-Jawzı̄,
Rada & Neuhäuser 2015). Ibn al-Athı̄r, and Ibn Abı̄ Zarc (Goldstein 1965),
– najm, which means just star, e.g. SN 1006 by al-Yamānı̄
According to historic observations and follow-up obser- and Ibn al-Daybac (Rada & Neuhäuser 2015),
vations, SN 1006 and its SNR G327.6+14.6 have a distance – nayzak, which can mean a comet or new star, e.g. SN
of 2.18 ± 0.08 kpc with very small extinction (Winkler et al. 1006 by c Alı̄ ibn Rid.wān and Ibn Abı̄ Zarc (Goldstein
2003); several arguments speak for a SN type Ia explosion 1965), but also something like spectacle or transient ce-
(see Schaefer 1996); for a SN type Ia, the peak apparent lestial event,
– athar, which means trace, but which was also used for
? Corresponding author: e-mail: rne@astro.uni-jena.de SN 1006 and SN 1054, and

c 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
790 Neuhäuser et al.: SN 1006 by Ibn Sı̄nā

– kawkab atharı̄ as spectacular star for SN 1054 by Ibn We would like to remark how this short text (Fig. 1)
Abı̄ Us.aybic a (Brecher et al. 1978). about what is most certainly SN 1006 was found: In his re-
If the observed object is classified just as some kind of view of Sezgin (1979), Goldstein (1982) reported that he
star, but the duration of appearance is limited (to e.g. a few (Goldstein) was told by A.I. Sabra that there is a mention of
months), the object can be identified as transient. The class SN 1006 in Ibn Sı̄nā’s Kitāb al-Shifā’:
of transient celestial objects is often characterized by fur- Professor A.I. Sabra informs me that a passage
ther details, whether e.g. star-like, stationary, and/or with in Avicenna’s Meteorology also mentions the super-
or without tail, which then classifies it as, e.g., nova, SN, nova of 1006.
or comet. See Kunitzsch (1995) for a review of the Arabic We present here the Arabic text from the edition of
words used for stars and transient celestial objects. Madkūr et al. (1965), page 73, lines 12 to 17 (see Fig. 1):
In their text book on historic SNe, Stephenson & Green
(2002) write in the chapter on Future Prospects: (line 12) fa-yac rid.u li-dhālika an yabqā
iltihābuhā wa-ishtic āluhā
In our view Arab writings have real potential as (line 13) muddatan .tawı̄la immā c alā s.ūrat dhu’āba
sources of further records of this SN [1006] – and aw dhanab, wa-aktharuhu shamālı̄ wa-qad yakūnu
possibly of that of AD 1054. janūbı̄yan, wa-immā c alā
Indeed, we present here such a new record of SN 1006. (line 14) s.ūrat kawkab min al-kawākib, ka-lladhı̄
We present the Arabic text and our English translation z.ahara fı̄ sanat sabc wa-tisc ı̄n wa-thalāth-mi’a
in Sect. 2. Then, in Sect. 3, we date the observation and li-l-hijra,
interpret the text. We summarize our findings in Sect. 4. (line 15) fa-baqiya qarı̄ban min thalāthat ashhur
yalt.ufu wa-yalt.ufu h.attā id.mah.alla, wa-kāna fı̄
ibtidā’ihi ilā l-sawād
2 Ibn Sı̄nā and his report about SN 1006
(line 16) wa-l-khud.ra, thumma jac ala kull waqt
Abū c Alı̄ al-H.usain b. c Abdallāh b. Sı̄nā (short Ibn Sı̄nā, yarmı̄ bi-l-sharar wa-yazdādu bayād.an wa-yalt.ufu
Lat. Avicenna) was a Persian polymath1 and lived from AD h.attā id.mah.alla, wa-qad
980-1037; he wrote books about theology, medicine, and (line 17) yakūnu c alā s.ūrat lih.ya, aw s.ūrat h.ayawān
natural sciences including astronomy; Ibn Sı̄nā follows in lahu qurūn, wa-c alā sā’ir al-s.uwar.
most topics Aristotle and Ptolemy, but also tried to improve Our English translation is as follows (words in round
on the quality and quantity of celestial measurements (see, brackets are missing in one or some manuscripts, square
e.g., Sezgin 1978). He invented the Jacob’s staff or cross brackets are our additions):
staff (Lat.: Baculus Jacobi) for precise altitude measure- It therefore happens that the burning and flaming
ments (Wiedemann 1927), later replaced by the sextant. In stays for a (long) while, either in form of a lock of
his works on the Almagest and in al-Shifā’, Ibn Sı̄nā de- hair or with a tail [i.e. in form of a comet], mostly in
scribes some of his own observations, including what he in- the north, but sometimes also in the south, or in form
terpreted as Venus transit,2 which was either a sunspot or the of a star among the stars [kawkab min al-kawākib]
Venus transit of AD 1032 May 24 (Goldstein 1969; Kapoor – like the one which appeared in the year 397(h).
2013). It remained for close to three months [qarı̄ban min
Ibn Sı̄nā’s encyclopaedic book entitled Kitāb al-Shifā’ thalāthat ashhur] getting fainter and fainter until it
(Book of Healing) is his major work on philosophy, written disappeared; at the beginning it was towards a dark-
from about AD 1013 to 1023; a nearly complete manuscript ness and greenness, then it began to throw out sparks
is located in the Bodleian Library, UK; a critical edition of [yarmı̄ bi-l-sharar] all the time, and then it became
the Arabic text has been published by Madkūr et al. (1965), more and more whitish and then became fainter and
which we have used for our work (see Fig. 1). In that work, disappeared. It can also have the form of a beard or
Ibn Sı̄nā discussed Aristotelian philosophy including natu- of an animal with horns or of other figures.
ral sciences. During the discussion of Aristotle’s Meteorol-
In the relevant chapter 5 of al-Shifā’, Ibn Sı̄nā discussed
ogy about transient celestial phenomena in the fifth volume,
the Meteorology of Aristotle. Following Aristotle, Ibn Sı̄nā
he mentioned a new star seen in 397h (AD 1006-1007).
explains that most atmospheric optical phenomena and in
1 Even though he was a Persian scholar, we think that it is correct to particular those connected with humid air would be due to
speak here of an Arabic report about SN 1006, because the transmitted text wet anathymiasis (evaporation), while all phenomena con-
itself is written in Arabic.
2 In book IX of his Almagest and very similar in the astronomy chapter nected with thunder, blizzard, wind as well as meteors and
of al-Shifā’: I say (that) I saw (myself) Venus as/like a black dot (spot) on comets (maybe including other transient celestial objects)
the surface/disc of the sun, given without date (Goldstein 1969; Kapoor would be due to dry anathymiasis. In the first sentence of
2013); this is confirmed by two later authors, namely by Nas.ı̄r al-Dı̄n al- the short quotation above, Ibn Sı̄nā obviously talks about
T.ūsı̄ (al Shaikh Abū c Alı̄ b. Sı̄nā mentions in his books that he had seen
what we today call comets (form of a lock of hair or with
Venus as a spot on the surface of the sun) and by Yahuda b. Solomon Ko-
hen (Avicenna saw Venus appearing like a spot in the midst of the sun) a tail). After describing the transient star of 397h, he con-
(Goldstein 1969). tinues to talk about other transient objects including what

c 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.an-journal.org
Astron. Nachr. / AN (2011) 791

Fig. 1 Here we show the Arabic text from the report of SN 1006 of Ibn Sı̄nā in al-Shifā’ from the Arabic edition by
Madkūr et al. (1965), page 73. The relevant text starts in the middle of the second line from the top and ends almost at
the (leftmost) end of the 3rd-to-last line from the bottom of the main text. The writing in the left margin is the Arabic line
number 15. The 4th line (line 14) reads (starting from the right) for the 2nd to 4th word kawkab min al-kawākib, i.e. a star
among the stars, and at the end of that line it specifies the year (the leftmost word is hijra). The lines at the bottom indicate
variant readings in different manuscripts, none of which change the content and meaning of the relevant text about the new
star: the words for long and hijra are missing in one or two manuscripts.

we call comets (it can also have the form of a beard ...). He menistan) at 37◦ 400 north to Kāth (now Uzbekistan) at
says that such an object stays for a (long) while, i.e. that it 41◦ 410 north, the ancient capital of the province of Kho-
is transient. That the new star among the stars is discussed rasan south of the Aral lake, which is now called Beruni in
together with what we call comets is not surprising, as both honor or the Arabic scholar al-Bı̄rūnı̄, who was born here.
refer to variable phenomena placed in the sub-lunar sphere. Ibn Sı̄nā left Kāth in AD 1012. Hence, if Ibn Sı̄nā was an
In other words, the term comet was in former times used for eyewitness of the new star in AD 1006/7, he would have ob-
several kinds of transient objects including what we today served it most likely from a location as far north as 36◦ 130 to
call comets, novae, and supernovae. 41◦ 410 , probably the latter. The text does not unfold whether
Since Ibn Sı̄nā could have been an eyewitness of SN Ibn Sı̄nā was the observer himself.
1006, let us consider where he was living, when SN 1006
was visible: Ibn Sı̄nā left Bukhara (now Uzbekistan) be-
tween AD 999 and 1005 and went via Nishapur (Iran) at a
geographic latitude of 36◦ 130 north and Merv (now Turk-

www.an-journal.org
c 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
792 Neuhäuser et al.: SN 1006 by Ibn Sı̄nā

3 Interpretation As mentioned before, all other observers gave much ear-

lier dates for their first observation: c Alı̄ ibn Rid.wān ob-
Let us first consider the dating of the observation, then the served SN 1006 since AD 1006 Apr 30, the Chinese ob-
other information content. servers since May 1 (but possibly already on Apr 3), and
in Japan, it was sighted first on Apr 28 or 30 (see Gold-
stein 1965; SG02); it is possible that SN 1006 was al-
3.1 Dating ready observed on around Apr 17/18 in Yemen, see Rada
& Neuhäuser (2015) for the evidence. The observer in St.
An apparent difference to other reports is the year of appear-
Gallen reports to have observed the new star for three
ance mentioned here by Ibn Sı̄nā, namely the year 397h,
months (SG02); if that was all after sunset, given his lo-
while the other Arabic reports all give 396h.
cation, he cannot have observed it after about July 10, so
The Muslim calendar is a lunar calendar, where the that he probably started to observe SN 1006 in April 1006.
months (and years) start with the evening when a new cres- While most Arabic observers mention that they observed the
cent moon is seen (days run from evening to evening), see SN for some 2-4 months, the Moroccan report mentions six
Quran, Sura 2, 189. There are no leap months. Months can months as visibility period, which would be until after con-
usually last 29 or 30 days, since the synodic month lasts junction with the Sun (Goldstein 1965; SG02). The Chinese
29.26 to 29.80 days (29.53 days on average). If in history the have observed also the helical setting and rising of SN 1006
Muslim months had alternating 29 or 30 days, then, in any in AD 1007 (SG02). Hence, SN 1006 was still visible in the
period of 30 years, there should be 11 years with one month Muslim lunar year that started on AD 1006 Sept 26 (±2) in
with one extra day (e.g. a lunar year with seven months with the evening. SN 1006 was in conjunction with the Sun from
30 days plus five months with 29 days). mid Sept to mid Nov, so if it was observed in 397h, then it
The year is given in the Muslim Hijra era, i.e. the num- must have been after mid Nov 1006.
ber of years after the start of the lunar year in which the There are four possibilities to be considered for the in-
Hijra took place, i.e. the emigration of the Islamic Prophet terpretation of the year (397h) given by Ibn Sı̄nā:
Muh.ammad from Mecca to Medina, known as Hijra. This
1. Ibn Sı̄nā as eye-witness mistakenly gave the slightly
era, i.e. the year 1h started on AD 622 Jul 16/17 according
wrong year in his text, e.g. a memory error or typo, pos-
to most scholars – but it may have been on AD 622 Jul 15/16
sibly years after the observation, al-Shifā’ was written
according to, e.g., de Blois (2000).
AD 1014-1020.
Any date given in the Muslim calendar can be converted 2. The year was changed by mistake during the (oral ?)
to a Julian or Gregorian date with a precision of ±2 days. transmission from the observer to Ibn Sı̄nā from 396h to
The reason for this uncertainty is, among the uncertainty 397h.
of the start of the era (see above), that it is not clear a- 3. A copying scribe made a mistake by changing the year
posteriori when in history a month had an extra day3 and from 396h to 397h.
that the first sighting of a new crescent moon can be de- 4. The observer (possibly Ibn Sı̄nā himself) in fact ob-
layed due to, e.g., bad weather and/or difficult landscape. served SN 1006 in 397h, which started on AD 1006 Sept
It is also possible that – even an experienced observer – 26 (±2), i.e. at or after heliacal rising after conjunction
claims to have seen the crescent new moon, even if it was with the Sun (after mid Sept).
not yet possible, so that a month would start one day too
early (Doggett & Schaefer 1994). See, e.g., Spuler & Mayr The latter (4) is less likely, because the star was brighter
(1961), de Blois (2000), Said et al. (1989), and Neuhäuser in the year 396h. The year 397h most certainly is given by
& Kunitzsch (2014) for more details about Muslim calendar mistake – just one year too late.
rules. There is no other transient celestial object that could
Since the date given by Ibn Sı̄nā is very rough, just the have been meant by Ibn Sı̄nā for 397h (AD 1006 or 1007),
year is given, we do not need to try to date the start of the in particular no comet (see Ho Peng Yoke 1962 and Kronk
year with high precision. According to the calculated Is- 1999).4 Further circumstantial evidence for a visibility pe-
lamic calendar (Spuler & Mayr 1961), the year 397h started riod in spring or mid AD 1006 is given by the fact that Ibn
on AD 1006 Sept 26 (±2) in the evening. 4 Cook (1999) quoted some additional text from Yahyā ibn Sac ı̄d al-
.
Ant.ākı̄ after the report for SN 1006 on what is most certainly a differ-
3 The artificially constructed calculated Islamic calendar uses leap days ent transient object, e.g. a meteor or bolide: Another star appeared with a
in certain, pre-defined years and months; in reality, we have to expect that, strong light in the west during the time of the falling of night during Satur-
in each period of 30 years, there were 11 months which had an additional day night, 9 Shawwal [10 July], and it stayed long and grew great. Then it
day (due to real, late crescent sighting) – in addition to those 354 days in broke up into three parts and disappeared. Appearance of the object in the
a pure lunar calendar year with – on average – six months of 29 days plus west at falling of night is not consistent with SN 1006, which was within
– on average – six months of 30 days (the average synodic month length is an azimut 1h of the meridian at sunset. Even if the text here says it stayed
29.53 days, and not 29.50). Due to late crescent sighting, the month with long and grew great, this can be a bolide, which stayed relatively long for a
one extra day did not necessarily follow the leap day/month rule in the bolide or meteor. The source for the latter object (bolide) can very well be a
calculated Islamic calendar used in, e.g., Spuler & Mayr (1961). Hence, different source than for the SN report. SG02 did not even include this extra
this calendar can deviate by up to two days (Ginzel 1906; Spuler & Mayr text in their citation of Yah.yā ibn Sac ı̄d al-Ant.ākı̄, obviously also because
1961; de Blois 2000). they consider it unrelated to SN 1006. Alternatively, the last sentences, if

c 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.an-journal.org
Astron. Nachr. / AN (2011) 793

Sı̄nā specified that the object was visible for close to three 37◦ 400 north, or Kāth 41◦ 410 north, probably the latter), he
months, which is quite consistent with the Arabic observers, could not observe SN 1006 for a long period (if he was the
who detected the object in April or May and monitored it for observer himself at all): His likely location Kāth has a sim-
typically two to four months. ilar geographic latitude as Naples, Italy (41◦ north), from
where a new star is related to a 3 month period: A very bril-
3.2 Stationarity, appearance, direction, duration, liant star shone, and a large drought happened for three
lightcurve, colour, and brightness months (SG02 from the Annales Beneventani). St. Gallen is
even further north at 47◦ 250 , and surrounded by high moun-
We can now discuss the other information content from Ibn tains, where the new star was also seen for 3 months (SG02).
Sı̄nā and compare it to other observers. As for the observer(s) in Naples, Ibn Sı̄nā could have ob-
Taillessness (and/or stationarity). With the wording in served SN 1006 only until about the end of July (if he ob-
form of a star among the stars, Ibn Sı̄nā probably means served only after sunset): On 1006 Jul 31, SN 1006 would
that the transient new object was tail-less – in contrast to the have been at an altitude of about 5◦ 300 above the horizon at
more common transient objects, comets with tails (if form of sunset with an apparent magnitude of about −1 mag. (We
a lock of hair/beard), which move relative to the stars. The also take into account that the southern horizon is quite flat
wording a star among the stars may also or alternatively as seen from Kāth/Beruni towards the south: Kāth/Beruni
mean stationarity. Other Arabic records mentioned the sta- is located at today’s border of Uzbekistan to Turkmenistan
tionarity: c Alı̄ ibn Rid.wān (It remained where it was and and the latter has almost no high mountains, in particular not
it moved daily with its zodiacal sign), Ibn al-Jawzı̄ wrote ... south of Kāth/Beruni.) Hence, the observer (whose report is
and it remained fixed ... (Goldstein 1965; SG02), and maybe transmitted by Ibn Sı̄nā) could have observed SN 1006 in
also al-Yamānı̄ and Ibn al-Daybac (remained unchanged). May, June, and July (not later, but possibly earlier in April).
Direction. Even though SN 1006 indeed appeared in the He may have observed (part of) the rise, the peak, the de-
far south as seen from Arabia or Persia for Ibn Sı̄nā, we can- crease, and even the heliacal setting of SN 1006. The fact
not conclude on the direction from his text. Even when he that Ibn Sı̄nā reports to have seen the new star for close to
says that [comets appear] mostly in the north, but sometimes three months (probably May, June, July) is fully consistent
also in the south, or in form of a star among the stars like with the other Arabic reports.5
the one which appeared in the year 397h, he may just quote Light curve. Ibn Sı̄nā also describes that the new
Aristotle for (normal) comets (or, more generally, transient star decreased in brightness before it disappeared (getting
celestial objects) to appear in both the north and the south, fainter and fainter until it disappeared and later and then
before then starting to discuss the new star of AD 1006. became fainter and disappeared), as do the Yemeni authors:
Duration. The duration of visibility given (qarı̄ban min Al-Yamānı̄ (In the night of mid-Ramad.ān, its light started
thalāthat ashhur for close to three months) can mean a little to decrease and gradually faded away) and similarly Ibn al-
less or a little more than three months, and it is consistent Daybac (its light diminished and it gradually faded away).
with most other observers: c Alı̄ ibn Rid.wān (four months), The connection of a gradual decrease in brightness with dis-
Ibn al-Jawzı̄ and Ibn al-Athı̄r (beginning of Shac bān ... until appearance by Ibn Sı̄nā could well mean the process of he-
the middle of Dhū al-Qac dah, i.e. 3.5 months), Moroccan liacal setting. Because SN 1006 was observed by some ob-
report (This star stayed for six months), Yah.yā ibn Sac ı̄d al- servers even after conjunction with the Sun since the end of
Ant.ākı̄ (it continued four months, Cook 1999), al-Yamānı̄ AD 1006, it is less likely that SN 1006 was not observable
(On the night of mid-Rajab in the year 396h, a star appeared any more due to intrinsic faintness before heliacal setting.
... In the night of mid-Ramad.ān, its light started to decrease The Arabic word fa-baqiya in the text by Ibn Sı̄nā was trans-
and gradually faded away, i.e. more than two months), and lated here as It remained (for close to three months), which
similarly Ibn al-Daybac (on the night of mid-Rajab a star probably does not mean It remained (fixed at its location),
like Venus appeared ... It remained unchanged until the but rather in a temporal sense (that it was seen for close
night of mid-Ramad.ān, i.e. not less than two months). In to three months), or possibly that it remained (somewhat
particular if Ibn Sı̄nā (or his source) observed SN 1006 since constant in brightness); also, stationarity may have been al-
about late April or early May 1006, and then for close to ready mentioned by Ibn Sı̄nā in his previous sentence (a star
three months, then he could have observed until heliacal set- among the stars).
ting for his location. Colour. The colour and its evolution is mentioned: at
Given that Ibn Sı̄nā was located quite far north in AD the beginning it was towards a darkness and greenness, then
1006, see Sect. 2, (either in Nishapur 36◦ 130 north or Merv it began to throw out sparks [yarmı̄ bi-l-sharar], and then it
became more and more whitish .... This part of the text may
pertaining to SN 1006, could either confirm the long visibility period and be more difficult to understand. What we translated as dark-
strong brightness (until July) or might be mis-dated (first appearance was
April/May instead of July as given here). The wording it broke up into three 5 For the observer in St. Gallen, the situation was even worse: he could

parts and disappeared might be understandable as effect of strong scintil- observe SN 1006 only until July 10 at most given his even more northern
lation at very low altitude, possibly around heliacal setting. When the sun location and the high mountains towards his southern horizon; since he
was some 9 − 18◦ below the horizon, SN 1006 was clearly visible in the observed for three months, he must have seen SN 1006 for his first time
west (azimuth being 1.5 to 2.5h west of meridian). before May.

www.an-journal.org
c 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
794 Neuhäuser et al.: SN 1006 by Ibn Sı̄nā

ness could even mean blackness or possibly faintness, but then it gradually decreased in brightness (probably due to
black is almost impossible as colour for a celestial object; heliacal setting).
while the meaning of black as unfortunate was also known In general, Ibn Sı̄nā’s text is consistent with other Arabic
and used in Arabic, this interpretation is unlikely here, be- (and non-Arabic) records about SN 1006, the main addition
cause it is combined with greenness – and also because is the colour evolution and some terms, e.g. star among the
Ibn Sı̄nā is known to have opposed black magic, astrology stars and yarmı̄ bi-l-sharar for throw out sparks. Hence, the
etc. The wording darkness and greenness could mean faint report is independent.
green-to-yellow (at the beginning, i.e. before peak bright- When c Alı̄ ibn Rid.wān told us that other scholars from
ness). What is described as green in the sky is often yellow time to time have followed it [SN 1006] and came to a simi-
or yellowish-to-greenish, as yellow and green are very sim- lar conclusion (SG02), he may have meant Ibn Sı̄nā, among
ilar to each other. The Chinese have reported that the new others.
star was yellow (SG02). Ibn Sı̄nā continues with then it be-
Acknowledgements. We thank the Institut für Geschichte der
came more and more whitish. In this case, Ibn Sı̄nā would
Arabisch-Islamischen Wissenschaften, Frankfurt, where we could
report that the new star was first yellowish-to-greenish and use the extensive library with Ibn Sı̄nā’s al-Shifā’ in Arabic. We
then more and more whitish. also acknowledge Dagmar L. Neuhäuser for various important
While it is also not clear whether our translation of the comments.
colours shows the intended meaning or even whether the
transmitted text is somewhat corrupt in this part, his text
can be interpreted in a consistent way as follows:
References
At the beginning it was towards a faintness and greenness Brecher, B.R., Lieber, E., Lieber, A.E. 1978, Nature, 273, 728
meaning that it was faint and greenish-yellowish at the be- Cook, D. 1999 JHA, 30, 131
ginning (like the Chinese report: it increased in brightness de Blois, F.C. 2000, Tārı̄kh, in: Bosworth, C.E., van Donzel, E.,
and was yellow, see SG02), then it began to throw out sparks Heinrichs, W.P., Lecomte, G. (Eds.) Encyclopaedia of Islam,
all the time, and then it became more and more whitish, i.e. new edition, Vol. X, Leiden: Brill
that it was scintillating during the period of largest bright- Doggett, L.E. & Schaefer, B.E. 1994, Icarus, 107, 388
Ginzel, F.K. 1906, Handbuch der mathematischen und technischen
ness (similar in China and in other Arabic reports) being
Chronologie I, Hinrichsche Leipzig
more and more whitish (brighter?) during peak brightness. Goldstein, B.R. 1965, AJ, 70, 105
The colour evolution reported is independent from other Goldstein, B.R. 1969, Centaurus, 14, 52
Arabic reports, both regarding the content and details and Goldstein, B.R. 1982, Isis, 73, 311
regarding the Arabic wording. Ho Peng Yoke 1962, Vistas, 5, 172
Brightness. Ibn Sı̄nā reports that it throws out sparks, Kapoor, R.C. 2013, Indian J. of History of Science, 48, 405
Kronk, G.W. 1999, Cometography, Vol. 1, Cambridge Univ. Press,
i.e. that it scintillated, this is again quite consistent with
Cambridge, UK
the other Arabic reports, consistent with a very bright lu- Kunitzsch, P. 1995, al-Nudjūm, in: Bosworth, C.E., van Donzel,
minosity: c Alı̄ ibn Rid.wān (it twinkled very much), Ibn al- E., Heinrichs, W.P., Lecomte, G. (Eds.) Encyclopaedia of Is-
Jawzı̄ (It was glittering), Ibn Sac ı̄d al-Ant.ākı̄ (It had daz- lam, new edition, Vol. VIII, Leiden: Brill
zling rays and a great rippling), as well as al-Yamānı̄ and Madkūr, I., Muntas.ir, c A., Zāyid, S., Ismāc ı̄l, c A. (Eds.) 1965, Ibn
Ibn al-Daybac (It showed a great turbulence). The word- Sı̄nā, al-Shifā’, al-T.abı̄c ı̄yāt, Vol. 5, Cairo
ing yarmı̄ bi-l-sharar for throw out sparks for scintillation Neuhäuser, R. & Kunitzsch, P. 2014, AN, 335, 968
is different from other Arabic reports, again showing its in- Rada, W. & Neuhäuser, R. 2015, AN, 336, 249
Said, S.S., Stephenson, F.R., Rada, W. 1989, Records of solar
dependence. Since Ibn Sı̄nā does not compared the new star
eclipses in Arabic chronicles, in: Bull. School of Oriental and
with either the brightest star(s) in the sky nor with Venus
African Studies, U London, Vol. LII Part I
(the brightest planet in the sky), the new star was proba- Schaefer, B.E. 1996, ApJ, 459, 438
bly much brighter than the brightest stars – and even much Sezgin, F. 1978, Geschichte des arabischen Schrifttums Vol. VI,
brighter than Venus, i.e. brighter than about −5 mag, the Astronomie, Brill Leiden
largest possible brightness of Venus. Sezgin, F. 1979, Geschichte des arabischen Schrifttums Vol. VII,
Astrologie und Meteorologie, Brill Leiden
Spuler, B. 1963, Der Islam. Zeitschrift für Geschichte und Kultur
4 Summary des Islamischen Orients 38, 154
Spuler, B. & Mayr, J. 1961, Wüstenfeld-Mahler’sche
The presented record by Ibn Sı̄nā about a transient celestial Vergleichungs-Tabellen, Dt. Morgenländische Ges., Steiner
object in 397h is quite clearly related to other credible re- Wiesbaden
ports about SN 1006, but original (however, dated one year Stephenson, F.R. & Green, D.A. 2002, Historical Supernovae and
Their Remnants, Oxford, Clarendon (SG02)
too late). Within a discussion about presumably sub-lunar
Stephenson, F.R., Clark, D.H., Crawford, D.F. 1977, MNRAS,
phenomena, Ibn Sı̄nā reports a transient celestial object in 180, 567
form of a tail-less star among the stars, which was seen for Tornberg, C.J. 1843, Annales Regum Mauretaniae, Vol. IX, Up-
close to three months, and it was scintillating; its colour may saliae
have changed first from greenish(-yellowish) to whitish, and Wiedemann, E. 1927, Acta Orientalia, 5, 82

c 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.an-journal.org
Astron. Nachr. / AN (2011) 795

Winkler, P.F., Gupta, G., Long, K.S. 2003, ApJ, 585, 324

www.an-journal.org
c 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim