Applied and Computational Historical Astronomy. Angewandte und computergestützte historische Astronomie.: Proceedings of the Splinter Meeting in the Astronomische Gesellschaft, Sept. 25, 2020. Nuncius Hamburgensis - Beiträge zur Geschichte der Naturwissenschaften; Vol. 55

By Gudrun Wolfschmidt

'Computational History' derives history from data and nowadays, therefore, relies on the technologies of the digital humanities. 'Computational History of Science' addresses questions of history by evaluating historical data, e.g. for tracing back copying traditions and conclude on transfer and transformation of data and knowledge. The term 'Applied Historical Astronomy', in contrast, tries to address questions of contemporary science by evaluating historical data in comparison with most recent data. This opens new possibilities, e.g. in the search for stellar transients among historical data.

In the contribution by Hoffmann & Vogt we will focus on the stellar transients among all the topics mentioned above. Philipp Protte discusses the accuracy of magnitudes and positions in ancient star catalogues, Andreas Schrimpf & Frank Verbunt present an analysis of an early modern star catalogue. Victor Reijs analyses the visibility of celestial objects for naked-eye observers, and Björn Kunzmann showcases some important variable stars in the history of astronomy. Rene Hudec presents astronomical photographic archives as a valuable data source for modern astrophysics. José M. Vaquero discusses the studies on solar observations made during the last four centuries. More technical are the contributions of Georg Zotti on Stellarium and Karsten Markus-Schnabel on data-mining and data-processing technologies. Ido Yavetz & Luca Beisel are developing a digital tool of computational history of science for the simulation of pre-modern astronomical models. Gerd Graßhoff focuses more on the application of computational history with regard to Kepler's Astronomia Nova while Tim Karberg presents an analysis of the astronomical orientation of buildings in the North Sudan.

• Language: English
• Publisher: tredition
• Release date: Aug 26, 2021
• ISBN: 9783347271067

Book preview

Introduction to Applied and Computational Historical Astronomy

Susanne M. Hoffmann (Jena) & Gudrun Wolfschmidt (Hamburg)

Abstract

‘Computational History of Science’ addresses questions of history by evaluating historical data, e. g. for tracing back copying traditions and conclude on transfer and transformation of data and knowledge. Modern computational history, of course, relies on the technologies of the digital humanities. The term ‘Applied Historical Astronomy’, in contrast, tries to address questions of contemporary science by evaluating historical data in comparison with most recent data. This term was introduced in the 20th century, but the method behind the term is much older. Here we give an introduction to the topic that is based on a review by Steele (2004) and enriched by our observations in the subsequent years. This way, we explain why and how the topic is still interesting. Basically the introduction of new methods of computational science developed in the recent years opens new possibilities. This introduction presents successes and open questions in the search for stellar transients among historical data.

Zusammenfassung: Einführung in die angewandte und computergestützte historische Astronomie

„Computational History of Science befasst sich mit Fragen der Geschichte, indem sie historische Daten auswertet, um z. B. Kopiervorgänge zurückzuverfolgen und auf Transfer und Transformation von Daten und Wissen zu schließen. Moderne Computational History stützt sich dabei auf die Technologien der Digital Humanities. Der Begriff „Applied Historical Astronomy hingegen versucht, Fragen der zeitgenössischen Wissenschaft zu adressieren, indem sie historische Daten im Vergleich zu neuesten Daten auswertet. Der Begriff wurde zwar im 20. Jahrhundert eingeführt; die Methode, die dahinter steckt, ist viel älter. Wir fassen hier einen Review von Steele (2004) zusammen, ergänzen diesen um weitere Beobachtungen und erklären, warum das Thema heute wieder aktuell ist. Dies liegt hauptsächlich an den neuen Methoden, die sich nach der rasanten Entwicklung der Computer-Technologie in den vergangenen Jahren ergeben. Diese Einführung präsentiert Erfolge und offene Fragen hinsichtlich der Suche nach stellaren Transienten in historischen Datensätzen.

1.1 Introduction to Computational History of Astronomy

1.1.1 Digitial History

Digital Humanities are an emerging research field which uses and develops new tools to address old and new questions. Modern electronic computer technology opens amazing possibilities and enables data-driven approaches for humanities. In 2016, there was an international workshop (Božić et al., 2016) on this topic in Dublin, dealing with computational methods like visualisation (Zwaan et al., 2016), data mining, database management and data curation (Feeney, 2016), artificial intelligence (AI) and natural language processing (NLP). These methods are applied to cultural studies such as investigations on rituals (Whitehouse, 2016), studies of distributions of historical epidemics (Marvel et al., 2013; Colavizza, 2016), literature studies (Grayson et al., 2016) and many more. The authors present tools to extract social networks of characters in 19th-century’s novels and the influence of social networks and number of travelers on the distribution of the Black Death in the 14th century as well as statistical analyses of human arousal in rituals depending on the ritual’s frequency. Even earlier, in 2012, a paper in the Journal of Statistical Mechanics promised to ‘bridge the gap between hard sciences and humanities’ with a ‘quantitative approach to evolution of music and philosophy’ (Vieira et al., 2012) and more recently, Mullen et al. (2019) suggests to map ‘the uncertainty of 19th century West African slave origins using a ‘Markov decision process model’.

Archaeology has always been in the intersection of (geo)science and (historical) humanities but in this millennium evolves to digital levels which is visible, for example in the name of the Austrian Ludwig Boltzmann Institute for Archaeological Prospection and Virtual Archaeology¹ and the German Max Planck Institute’s research project VII Digital and Computational History of Science.² These are only a few examples to sketch the broad range of applications from mathematics and computer science for classical questions of humanities.

In general, digital humanities provide new approaches to the old questions and put humanities on a profound base of data instead of mere ‘impressions’ and ‘reading between the lines’. The base of this data-driven historical science are extensive data archives that are currently created all over the world by research in digital humanities. All museums and libraries pursue digitalization projects of their books, manuscripts and 3D-objects (using 3D-scanning). Digital collections are created like, for example, the digital museum Europeana,³ the search tool for archaeological objects Arachne,⁴ the digital cuneiform library CDLI,⁵ the repository for the history of ancient and prehistoric science by Edition TOPOI⁶ and the database for global history Seshat:⁷ see Fig. 1.2. Currently huge editions are published online, machine readable, searchable and open access, eg. Ptolemaeus Arabus et Latinus⁸ of the Bavararian Academy of Science or the edition humboldt digital⁹ of the Berlin-Brandenburgische Akademie der Wissenschaften. These new digital editions perfectly qualify for all types of machine-learning analysis.

Additionally, modern journals allow data publication in addition to any paper. Free libraries and repositories to which everybody can contribute (e. g. wikipedia, wikimedia) and more sophisticated open repositories like Wikidata,¹⁰ Zenodo¹¹ and GitHub¹² can be combined with information in the classical media (e. g. books, documentary films). These new media and new methods enable the researcher to access and consider much more information at once.

There are tools for researchers like The History Playground¹³ for discovering temporal trends in massive text corpora by Bristol University (Lansdall-Welfare & Cristianini, 2018, and many references therein) as well as tools for learners like the European Time Machine¹⁴ project ‘creating a collective digital information system mapping the European economic, social, cultural and geographical evolution across times’.

Summarizing, digital history cares for the setup of digital archives, develops text parsers and text mining tools with the according statistical analysis and its visualization (that has always been performed in history), deals with the analytics of cultures and macroscopic trends in cultural change, online publishing and data mining projects in all media repositories.

Figure 1.2:

Screenshots of two huge databases dedicated especially to historical topics. These are only two of many examples for digital historical data publicly available

The resulting challenge to handle these data is, of course, also addressed with computational means: Visualisation tools that allow real-time requests on the continuously growing on-line repositories and databases have the potential to make all results always up to date and as reliable as the basic data are: The epoch of long lasting research results is definitely over. Instead, long lasting are only the display tools and methods while the continuously added data continuously changes the output.

However, the great advantage of this digital research is that large sets of historical documents can be analyzed in a comprehensive form and phylogenetic interdependencies, the transfer of manuscripts, the diffusion of ideas and technological innovation processes can easily be visualized and added to the global knowledge. In general history, this is sometimes called an ‘Empirical Turn’ in humanities.

Of course, this new technology comes with a huge variety of criticism because of the challenge for the researcher to correctly treat and interpret the data. Classical studies of academic subjects in humanities do not include courses on the evaluation of data, error bars and reliability of data. These important topics have been a key issue in science since centuries but are hardly addressed in humanities. Therefore, untrained people in research or in the public show – of course – a healthy scepticism with regard to digital humanities. They point out the ‘black boxes’, unreflected usage of tools by many computer users and the lack of ability of peers in a scientific peer-reviewing process.¹⁵

However, I think, this is only a critique on incompetent users and does not at all affect the methods themselves: A good method does not turn bad in case some or many users do not understand it. Instead, the criticism should be read as an appeal to train the users appropriately.

1.1.2 Computational History of Astronomy

More specifically for the history of science, the new computational research group at the Max Planck Institute for the History of Science states on their website:¹⁶ ‘Computational History of Science uses algorithmic methods to solve novel, challenging questions in the history of science. These methods enable us to execute, document, and communicate sophisticated investigations.’ The electronic tools might be new but the computational methods itself have been common in the history of (natural) science since long time.

In the history of astronomy in particular, computational methods have always been used: for example, Hipparchus’ historic star catalogue from the –2nd century is not directly preserved but fragments of it remain in the introduction to the Almagest star catalogue and in a commentary on Aratus written by Hipparchus’ own hand. Thus, since centuries, scholars in astronomy and in the history of science try to reconstruct the lost catalogue by computations from the preserved fragments, e.g. Vogt (1925), Graßhoff (1990), Hoffmann (2017). Graßhoff (1990) analyzed Ptolemy’s star catalogue with statistical means and used correlation plots to visualise the long-known dependencies of Ptolemy’s and Hipparchus’s data. A recent study Cardozo Dias & Stuchi (2013) shows that the use of numerical computations can contribute to the understanding of Isaac Newton’s paths of generating knowledge on gravity while the contribution by Graßhoff & Abkenar in this book, chapter 6, p. 91, performs an analysis of Kepler’s Astronomia Nova in order to address the question on the origin of his data (observed or computed).

Since decades, astronomers and historians of science compute and use tables of historical eclipses and solar system data, e. g. Tuckerman (1964), Stephenson & Houlden (1981), Espenak & Meeus (2006). Through the development of the World Wide Web as a global internetwork of computers and servers, these tables are globally accessible. Everybody can use them – scientists for questions in modern science and historians for questions on history. This way, historians can more easily address questions on dependencies of historical data from each other, e. g. Graßhoff (1990), Hoffmann (2017).

Modern computational history of science, of course, uses the methods of digital humanities: In principle, the programming language and code editors that are used are not a matter of scientific discussion because knowledge is independent of the tool that is used to denote or generate it. However, Wolfram Mathematica and its ‘computable document format’ (CDF) seems to have been used for a long time while more recently, Jupyter notebooks turn out as a viral environment for the development and as an appropriate tool for simultaneous publications of a book and its underlying data as they can be designed as executable documents. Jupyter notebooks serve as environment for the systematic analysis of data of all types (numbers, text, pictures, video and audio) by computational treatment with Python scripts.

Additionally, there are digitalization projects dedicated especially to create databases for historical astronomy. For instance, the art historian Saxl Project¹⁷ collects illustrated manuscripts of astral science from the medieval and Renaissance epochs, while the DISHAS database project¹⁸ assembles post-hellenistic pre-modern data in tabular form and, thus, collaborates with the research projects on the edition of historical astronomical tables from the Alfonsine Europe, India and the Ptolemaeus Arabus et Latinus-project to collect the literature.

Figure 1.3:

Screenshots of the British Saxl Project and the French DISHAS project, both dedicated to the history of astronomy and astrology

Summarising, topics of Computational History of Astronomy concern everything that is computable and can be analysed on the basis of data. With the development of multimedia-computers in the last decades of the 20th century, more and more data of humanities become accessible with computational means: Modern analytical 3D scanners, software tools and programming languages are able to generate and treat 3D models of historical sculptures and objects as well as 2D pictures and 1D text.

Digital and data-driven history of astronomy, of course, faces the same difficulties as general history in the age of digital humanities and big data. However, astronomers are much more used to work with databases than other disciplines: The stellar database SIMBAD and the according ‘catalogue of catalogues’ VisieR of astronomical ground based and satellite data have been available on-line at the Centre des Données astronomiques Strasbourg (CDS)¹⁹ since the 1990s. There is already a generation of professional astronomers who have been trained in data mining, data-reduction in invisible wavelengths, usage of virtual observatories and other computational tool rather than in the manual usage of telescopes. Digital Historical Astronomy should, thus, combine the two streams from history and inaccessible modern instruments (such as space telescopes and multi-messenger detectors).

While computational history of science addresses questions of history, e. g. how human society developed methods and knowledge or how methods and knowledge were transferred and transformed from one historical culture to the other, applied historical astronomy aims to apply this knowledge on questions of modern astrophysics.

1.2 Introduction to Applied Historical Astronomy

A query for the term ‘applied historical astronomy’ (AHA) in the NASA abstract server returns papers back to 1981. The review by Steele (2004) summarizes two slightly different definitions:

(i) the fact that historical data is used to address current research questions and

(ii) the usage of methods borrowed from the history of science.

With the second condition, he considers historical astronomy an interdisciplinary research field that combines methods from the history of science and astrophysics.

Focussing on cases where the used historical data had been produced in a different culture as the user belongs to, prevents Steele from having to discuss the fact that all research and especially most processes in space evolve on long time scales and research always have to rely on data that has been collected by other people.

1.2.1 Usage of historical data

Most processes in space evolve on large time scales with regard to human life span. Additionally, astronomical research always has to rely on data that has been collected by other people. Often, these other people are already dead when their observational data lead to new conclusions. The user of these data, thus, cannot ask the observer anymore about the technique and error bars and so on. Just to mention some of the most popularly known historical cases:

2000 years ago: precession. According to the Almagest, Hipparchus used observational data from Timocharis and Aristyllus who lived 100 to 150 years before himself. In comparison with Hipparchus’s own measurements, he proved the shift of stars due to precession. Ptolemy, another 260 years later, added observational data from several Roman observers and his own data to support this original proof. Thus, they both used data from observers that were already dead.

18th century: enlargement of the solar system. When William and Caroline Herschel discovered ‘George’s Star’ in 1781, they thought it was a comet. However, the object moves extraordinarily slowly. The lucky circumstances that the object was identified with an additional star among Flamsteed’s observations from 1690, made it possible to determine the orbital parameters of this object rather quickly and, thus, prove that this is a planet (now called Uranus). Flamsteed had died 1719, his star catalogue had been published in 1725.

19th century: size of the solar system. In 1660 the French astronomer Joseph-Nicolas Delisle suggested global observations of the Transits of Venus ahead (1661 and 1669) in order to perform Halley’s suggestion for the determination of the astronomical unit. Delisle himself died 1668 but his colleagues and students kept collecting data from expeditions all over the world. Yet, all results for the distance of the Sun remained preliminary until the student of Gauss, Johann Franz Encke (1791–1865), applied the newly developed mathematical methods of error calculations and least square fits to the data in 1820 to 1824. In his time, all observers of the 17th century and some of the observers of the 18th century were already dead, so he used historical data.

… These are only some of the historic highlights; in fact, most astronomers use historical data everyday, e. g. when we use the Bright Star Catalog that goes back to at least 1908 (continuously revised) and compare it to recent Gaia data. We are so used to use historical data that we do not even notice it in most cases.

It might be added that the Babylonian goal year method is a technique of almanach makers to predict the positions of planets by using old astronomical diaries. These historical diaries might be 8 years old (for Venus) but some decades for Jupiter and Saturn. This example shows that the method of working with data archives from the own astronomical culture is at least 2500 years old and one could ask whether or not the usage of archival data is already AHA.

With the aforementioned historic discoveries in mind, Steele’s definition of applied historical astronomy (AHA) as the usage of data from another culture, makes AHA more rare and more specific. This way, Steele (2004) discusses that the usage of Hipparchus’s star catalogue in Ptolemy’s Almagest is not AHA but the usage of Babylonian solar eclipse observations in the Almagest is. Hence, both aspects of the definition declare ‘applied historical astronomy’ as a beautiful modern designation for an old and very common method – at least going back to Greek antiquity.

1.2.2 The aspect of application

The chair of F. Richard Stephenson in Durham had the title ‘Applied Historical Astronomy’. On its website, it is explained that ‘Ancient records from China, Japan, Korea, Babylon, and Arabia, as well as European medieval sources, are studied. They provide long baseline data for phenomena such as the Earth’s rotation, the Solar cycle, comets, novae and supernovae. ’

This sentence perfectly describes the goal of this area in astronomy: We wish to make historical data usable for modern astrophysics. Stephenson himself is ‘widely recognized as founder of the specialist field of Applied Historical Astronomy’ as it is written in the Festschrift to his honor (Orchiston et al., 2015, p. xxviii).

As the topics of Stephenson’s research aim to apply historical data, of course he also studied historical maps and basics concepts of historical astronomy. Studying the production of data, historical measurements and the possibilities of naked-eye observations is one of the bases of this field.

Other scholars added the question on the long-term variability of stars. Fujiwara et al. (2004) studied the reliability of magnitudes in old star catalogues, Lequeux (2014) and Protte & Hoffmann (2020) contributed some error estimates and transformation equations for a few particular historical star catalogues and Fujiwara & Hirai (2019) presented a study of unknown stars in historical star catalogues. They concluded to have found 13 unknown objects in seven star catalogues and four maps.

Hamacher (2018a) studying the cultural traces of astronomical knowledge in the Aboriginal Australian culture, suggests knowledge on the variability of red giant stars could be encoded in some particular traditions. He also discusses naked-eye observations (Hamacher, 2018b), the Great Eruption of η Carinae in the 19th century (Hamacher & Frew, 2010), and the apparent lack of observations of supernovae in this astronomical culture (Hamacher, 2014).

1.2.3 Exploring stellar evolution: The needs of astrophysics

The need of this research is clearly visible: historically with regard to the explanation of observations of transients, today especially with regard to the research on close binary systems but there are a lot of difficulties.

1.2.3.1 New stars – an historical example

When Tycho Brahe (1572) and Johannes Kepler (1604) observed unexplained ‘new stars’ (stellae novae), they had no idea on the reasons for this phenomenon. Kepler (1606) even speculated that the close conjunction of planets in Ophiuchus in December 1603 somehow ‘produced’ a new star: In December 1603, Jupiter and Saturn had a conjunction with a closest approach of ~1°, and the day before Christmas (morning of Dec 23th) Mercury stood between them, separated from Saturn by only ~20arcmin. When the supernova came into being, Jupiter and Mars stood only ~2° and ~3.3° away and Saturn approached (passing by two months later, unobservable due to the vicinity of the Sun). Kepler’s speculations even led to the (wrong) hypotheses of the Star of Bethlehem was such a ‘star produced by planet conjunctions’ (Occhieppo, 2003) which is popularly narrated since decades (Letsch, 1953) although it is known that this should not be treated as astronomical problem (Söding et al., 2013; Barthel & van Kooten, 2014). Also in the 17th to 19th century, nobody had any idea of the reason for these stellar transients.

After the start of the usage of telescopes in astronomy (1610) and especially in the epoch of the huge celestial atlases when more stars²⁰ and even nebulae of various type (Jones, 1975; Latusseck & Hoffmann , 2017) were included into star catalogues, the number of observed transients increased rapidly: 4 ‘novae’ in the 17th century, one in the 18th, 16 in the 19th century. Looking for more data to substantiate this exploration, scholars started to collect ancient records of similar transient phenomena from chronicles of China and its colonies, cf. e.g. Biot (1843); Humboldt (1850).²¹ However, the reasons for transients were still speculation on the relevance of nebulae (e. g. that stars flare up as result of interaction with nebulae, a phenomenon similar to shooting stars but bigger). When a very bright (GK Per: 0 mag) new star appeared prominently (Archenhold, 1901) in February 1901, it was not even understood how stars produce their energy neither how they evolve or that many ‘nebulae’ are galaxies outside the Milky Way. Only the 20th century brought some advance in understanding the physics behind the phenomena.

Figure 1.4:

Henry / Heinrich III observes the new star over the city Tivoli / Tyburtina (Supernova 1054)

(Workshop of Diebold Lauber, unknown artist, around 1450, https://digi.ub.uni-heidelberg.de/diglit/cpg149/8823/text_heidicon)

1.2.3.2 Current research on novae

Example 1: In their abstract of the interesting paper(s) on BK Lyn, Patterson et al. (2013) write:

Reviewing all the star’s oddities, we speculate:

(a) BKLyn is the remnant of the probable nova on 101 December 30, and (b) it has been fading ever since, but it has taken ~ 2000 yr for the accretion rate to drop sufficiently to permit dwarf-nova eruptions. If such behaviour is common, it can explain other puzzles of CV evolution.

The suggestion to identify the transient 101 with a nova of the system BK Lyn dates back to 1986. The research of Kemp & Patterson et al. (2012) concerns the change of light curve shapes of this particular star. During a 20 yr-campaign, a clear transition from a nova-like to dwarf nova has been observed. This is a real breakthrough and confirms expectations on the evolution of close binaries. However, the speculation that it had been fading ever since the end of year 101 CE is a bit dangerous because dwarf novae of the system would not be detectable for the naked eye. This is probably the reason for the author to treat it as speculation.

Example 2: Shara et al. (2007) announced the discovery of a nebula in the vicinity of the prototype dwarf nova Z Cam. They write about the age estimate:

The Z Cam physical shell size then sets an age range of   2,400 – 240 yr, with the larger value favoured as significant snowploughing has almost certainly taken place. Records of erupting stars two or more millennia ago are almost non-existent, so it is not surprising that no historical record of a Z Cam nova eruption exists.

Later they continue

It strengthens the claim that cataclysmic variables can undergo cyclic, metamorphic transitions from one sub-type to another. It argues that the present-day, prototypical dwarf nova Z Cam underwent a classical nova eruption a few thousand years ago. During that eruption it must have become, for a few days or weeks, one of the brightest stars in the sky.

From this Nature paper we conclude the following:

• The age estimate in astrophysics often allows not even to give the order of magnitude. The nebula could be 240 or 2400 years old. Thus, an historical date of the observation of this eruption would clearly improve the model.

• It would be possible for humans to observe this because it made Z Cam a bright appearance.

• Historical records of the time ~ 2000 years ago do rarely exist – but they do exist and the margin of error for the age estimate allows to search in a huge temporal range.

In case, some observational examples of this type could be provided, this could enlarge the baseline of our knowledge on more than one binary system and allow some more general considerations on the claim that cataclysmic variables can undergo cyclic, metamorphic transitions from one sub-type to another.

1.2.4 Successful stories

Probably the most famous success of applied historical astronomy was achieved before the establishment of this term. It is the identification of the Chinese guest star in 1054/55 with the nebula known as M 1 as its remnant.

1.2.4.1 Supernovae

Around 1920, Rutherford developed the idea of nuclear fusion as the star’s source of energy and in the 1930s, when the neutron was discovered (Chadwick, 1932), the astronomers Walter Baade and Fritz Zwicky suggested to call the ‘new stars’ of really big amplitudes (> 19 mag) super-novae, cf. Baade & Zwicky (1934,b, 1938); Baade (1938), and they speculated that these objects might refer to explosions of (massive) stars whose remnants could be neutron stars, cf. Baade & Zwicky (1934c); Zwicky (1935, 1938, 1939). With this idea, the historical records of the enormously bright transients in 1054 (Baade, 1942; Mayall & Oort, 1942), 1572 (Baade, 1945) and 1604 (Schlier, 1935; Baade, 1943) were considered. There are many records preserved of SN1054, reporting an impressive sighting ‘southeast of Tianguan’ where the term Tianguan is normally identified with ζ Tau (Mayall & Oort, 1942). The nearby nebula M1 that had been known since the 18th century is not southeast of Tianguan but Tianguan is southeast of the nebula and a central star had been discovered by Baade (1942). Hence, this deep sky object was quickly connected to the historical record and – with some luck – some decades later, in the 1960s, the nebula’s central star turned out to be a pulsar (Bell & Hewish, 1967; Hewish et al., 1968) which confirms both, the brave idea of stars made out of neutrons as well as the also brave idea that the nebula could, thus, be related to the historical record. Detailed reviews on supernovae: Trimble (1982, 1983, 2007).

In the 1970s, the influencing work of Richard Stephenson started with a systematic search for similar cases in the collections of historical records: Stephenson (1976), Clark & Stephenson (1977), Stephenson & Green (2002), Green & Stephenson (2017). Carefully, he read and selected records that do not report a movement or tail and suggested those with a given duration of several months as supernova candidates. In the subsequent decades, Stephenson and his colleagues kept looking for supernova remnants as counterparts of these historical records. However, they correctly always state, that our knowledge of possible counterparts continuously increases and that every suggestion they (or anybody else) make, would be preliminary.

1.2.4.2 Novae

However, supernovae explained only a few of the historical sightings while many transients with smaller amplitude remained unexplained. Especially, the outburst of DQ Her in 1934 that reached 1.4 mag (as bright as Deneb in the popular summer triangle) kept puzzling scholars. Further lists of historical guest stars were made: Lundmark (1923), Hsi (1957); Ho (1962); Xi et al. (1966); Pskovskii (1972).

In the 1960s, it was realized that the novae that do not have ‘super’ amplitudes, originate from close binaries: Kraft (1962); Kraft et al. (1962); Kraft (1964a,b). This was the key for understanding they are permitted from cataclysmic binaries (CVs) which was beyond the scope of Stephenson’s systematic search. Occasionally scholars were interested in historical observations of novae (Payne-Gaposchkin, 1957; Duerbeck, 1987, 1992), discussed and rejected particular cases as historical novae (Shara et al., 2017b; Göttgens et al., 2019) or suggested identifications of CVs with historical records, e.g. Sabbadin & Bianchini (1983); Hertzog (1986); Ringwald & Naylor (1997); Robertson et al. (2000); Patterson et al. (2013), V529 Ori and BKLyn (Johansson, 2007), the historical sighting of –76 (Kamiński et al., 2015), CK Vul and the discoveries of nova shells misclassified as planetary nebulae, e. g. Miszalski et al. (2016), for object Te 11 (Shara et al., 2017a).

In our series of papers in 2019 to 2020, we suggested a more systematic method to deal with historical records: The era of digital humanities and data-mining in astrophysics provides new search