A Very Short History of the Humanities:

Patterns versus Interpretations

DRAFT Version August 2015

Rens Bod
University of Amsterdam
rens.bod@gmail.com
http://devergetenwetenschappen.blogspot.com

This paper sketches the long term history of the humanities by discussing the empirical versus the
interpretative traditions in the humanities. For several centuries, empiricism has almost exclusively
been attributed to the sciences. The sciences search for general patterns and principles, while the
humanities aim at understanding unique events. The sciences try to explain the world, while the
humanities aim at interpreting it. The sciences aim for objectivity while the humanities are subjective
and speculative. These oppositions are much older than C.P. Snow’s well-known two cultures debate;
they can be traced back to Antiquity and are – perhaps surprisingly -- rooted in the disciplines that
we nowadays call ‘humanities’. From at least the third century BCE onwards, there have been two
kinds of practices in philology, historiography, poetics, the study of art and the study of music. On the
one hand people searched for regularities and patterns, while on the other hand they searched for the
unique and the exceptional. Both approaches have not disappeared from the humanities since then. In
this essay I shall sketch the longue durée of the pattern-seeking tradition in the humanities and
compare it to the interpretative tradition. 1 I will argue that interpretations were not always in
opposition to pattern-seeking but were often constructed on the patterns found. Finally I will come
back to the relation between humanities and science, arguing that the common wisdom that the
humanities are moving towards science when they search for patterns is mistaken. The search for
patterns has been a continuous line in the humanities from Antiquity onwards.

Perhaps the oldest debate which contrasts patterns to interpretations may be found in
philology. This debate, which was first mentioned by the Latin grammarian M. Terentius
Varro (116-27 BCE), has come to be known as the controversy between the analogists of
Alexandria and the anomalists of Pergamon. 2 With the establishment of the library of
Alexandria hundreds of thousands of manuscripts – and remnants thereof -- were brought

1
This paper is partly based on my book A New History of the Humanities: The Search for Principles and
Patterns from Antiquity to the Present (Oxford University Press, 2013). As short as it is, the paper goes beyond
my book in that it deals not only with patterns (and principles) but also with interpretations and the relation
between patterns and interpretations. The paper could be seen as first, necessary extension of my book.
2
This controversy is nowadays taken as an artifact of Varro’s way of arranging material -- see Detlev Fehling,
“Varro und die grammatische Lehre von der Analogie und der Flexion”, Glotta, 35, 1956, pp. 214-270, and 36,
1957, pp. 48-100 (with thanks to Glenn Most, p.c.). Although no debate may actually have taken place between
analogists and anomalists, there were two contrasting practices in ancient philology.

1
together, representing an empirical world of texts never seen before. Among the often dozens
or even hundreds of copies of the same text, no two were alike. In some cases the differences
were modest and had come about because of copying errors, but the discrepancies could also
be substantial, consisting of whole sentences that appeared to be deliberate changes, additions
or omissions. And there were also texts that had only survived in the form of incomplete
fragments. How could the original text – the archetype – be deduced from all this material?

The first librarian of the Alexandrian library, Zenodotus of Ephesus (c.333-c.260

BCE) tried to tackle this problem by compiling a dictionary of typically Homeric words. This
way he hoped to be able to formulate the ‘perfect’ Homeric text from the many corrupt
remnants of manuscripts. 3 Most of Zenodotus’s criteria were however based on aesthetic
preferences. His successors Aristophanes of Byzantium (c.257-180 BCE) and Aristarchus of
Samothrace (c.216-c.144 BCE) wished to keep philology free from such subjective elements.
They opted for an explicit method to figure out how an unknown word form can be identified
either as an archaic word or as an error. Aristophanes approached this problem on the basis of
a concept of analogy. 4 If one could establish that an unknown word was conjugated or
declined in the same way as a known word, it could be taken as an archaic rather than as a
corrupted word. Aristophanes defined five criteria that word forms had to comply with
among themselves in order to be described as ‘analogous’. The word forms had to correspond
in regard to (1) gender, (2) case, (3) ending, (4) number of syllables and (5) stress (or sound).
The Alexandrian philologists used the designation of analogía for their search of analogical
patterns across word forms.

There was also a competing, Stoic school that was established in Pergamon by Zeno
of Citium (334-262 BCE). The philologists of this school searched for exceptions rather than
for regularities. They worked on the basis of what they called anomalía, which was
introduced by Chrysippus of Soli (c.280-c.207 BCE). Instead of looking for analogies
between word forms, Chrysippus stressed the importance of seeking exceptions. According to
the anomalists it was impossible to deduce the original form of a text on the grounds of
analogies. According to the most fervent supporter of this approach, Crates of Mallos (died
c.150 BCE), all the efforts expended by the analogists were vain and superficial. The
philologist’s task was to select the surviving document that came as close as possible to the

3
Klaus Nickau, Untersuchungen zur textkritischen Methode des Zenodotos von Ephesos, De Gruyter, 1977.
4
For an in-depth study on Aristophanes, see Christopher Callanan, Die Sprachbeschreibung bei Aristophanes
von Byzanz, Vandenhoeck und Ruprecht, 1987. See also Rudolf Pfeiffer, History of Classical Scholarship:
From the Beginnings to the End of the Hellenistic Age, Clarendon Press, 1968, pp. 202-208.

2
author’s intentions and, once this selection had been made, the chosen document had to be
adhered to as closely as possible and interpreted correspondingly. 5 According to the
anomalists no deeper system of regularities or patterns existed. This method was therefore
wilder than the method based on analogies, and often resulted in highly fanciful and
allegorical text interpretations. 6

Nevertheless, the anomalist approach produced a number of extraordinarily original

works. The anomalists – unlike the analogists, whose work was mostly formal – produced
erudite commentaries. For example Demetrius of Scepsis wrote a series of thirty books about
the Trojan forces, which were addressed in fewer than sixty-two lines in the entire Iliad.
Every point of view was dissected by the author, using a vast quantity of literature, local and
oral traditions, history, mythology, geography, poetry and observations by travellers – in
other words he called upon the entirety of classical knowledge to contribute to the
interpretation of the text. 7

Despite the anomalist critique on the analogist way of working, the formal
Alexandrian method has withstood the test of time as a critical approach to text
reconstruction. We owe a debt of gratitude to the insights of the Alexandrians for the editions
of Homer, Hesiod, Pindar, Archilochus and Anacreon, and of the tragedians and historians,
which have been handed down to us.

Thus the two schools – analogist versus anomalist – involved not just two different
methods, they had different goals too. The analogist search for empirical patterns served to
derive the original text, while the anomalist focus on the exceptional served to interpret a
specific version of the text. This opposition – searching for patterns versus understanding the
unique – seems to foreshadow the two apparently opposing ways of working that we also find
in today’s humanities disciplines: explaining versus understanding, the search for the general
versus the search for the exceptional.

There are similar oppositions in other ancient humanistic disciplines. In their

descriptions of the Persian and Peloponnesian Wars, the historians Herodotus (c.484–425
BCE) and Thucydides (c.460–c.395 BCE) believed they could recognize a cyclical pattern in

5
Maria Broggiato (ed.), Cratete di Mallo: I frammenti, Agorà Edizioni, 2001. See also David Greetham, Textual
Scholarship: An Introduction, Garland Publishing, 1994, p. 300.
6
Cf. John Edwin Sandys, A Short History of Classical Scholarship from the Sixth Century B.C. to Present Day,
Cambridge University Press, 1915, p. 49.
7
For fragments of Demetrius of Scepsis, see http://www.digitalclassicist.org/.

3
past events. Herodotus’s Histories, for instance, reflected a repeating pattern of rise, peak and
decline. We see this pattern in his descriptions of both people and states, for example the
tyrant Pisistratus and Athens, King Croesus and Lydia, and Darius and Persia: their fortunes
rose and fell. Herodotus considered the cyclical pattern to be the basic structure of history:
‘For many states that were once great have now become small, and in my lifetime those that
are great used to be small.’ 8 Thucydides also contended that the rise and fall of Athens and its
disintegration during the Peloponnesian Wars had parallels with other historical periods, and
believed that the cyclical pattern was analogous to human nature and therefore could even
serve as an ‘aid for interpreting the future’. 9

Contrary to Herodotus and Thucydides, the Greek historian Polybius (c.200–c.118

BCE) rejected the general cyclical pattern in history. Instead he glorified the unique and the
exceptional character of his new patria Rome. Polybius expressed great admiration for the
way Rome succeeded where the Greeks had failed. Rome, he argued, refuted the cyclical
pattern of rise, peak and decline that had occurred in the history of Athens, i.e. a cycle of
monarchy, aristocracy, oligarchy, democracy and via tyranny back to monarchy again. 10
Unlike Athens and other cities, Rome was immune to this cycle – and therefore to decline –
because of its mixed constitution. In Rome at the time of Polybius there was a monarchy (the
consuls), an aristocracy (the senate) and a democracy (the people’s assemblies) all at the
same time. According to Polybius this simultaneity broke the cyclical pattern, which turned
the history of Rome into a unique non-cyclical history, so he believed. 11

Although we cannot speak of a debate between Herodotus and Thucydides on the one
hand and Polybius at the other (as they lived more than two centuries apart), an opposition
between the two ideas is nevertheless discernable: where Herodotus and Thucydides saw a
recurring pattern in the history of different states and peoples, Polybius saw the uniqueness in
the history of Rome. It is interesting that a similar opposition between patterns and the unique
is found in the work of Chinese historians during the Han period: where the imperial scribe
Sima Qian (c.145-86 BCE) recognized a cyclical pattern of rise and fall of Chinese dynasties,

8
Herodotus, Histories, 1.5.
9
Thucydides, History of the Peloponnesian War, 1.22.
10
Polybius, Histories, 1.1-2.
11
Polybius, ibidem, 1.4.

4
the historian Ban Gu (32-92 CE) believed that the dynastic family's mandate of the western
Han dynasty was exceptional and permanently bestowed. 12

This opposition between searching for the regular and searching for the exceptional is
also present in the study of literature, art and music in Antiquity. For example, while Aristotle
and others found regularities and rules for ‘good’ narratives, 13 Longinus argued that
‘sublime’ narratives do not follow any rules but, instead, are exceptional. 14 And while Pliny
found a pattern in Greek and Roman art, which could be defined by mathematical proportions
known as the canon, 15 the same Pliny contended that ‘beautiful’ art depended on inspiration
or even on coincidence. 16 And while the ‘first musicologist’ 17 Aristoxenus found rules for
harmony and melody in Greek musical pieces, the Harmonists opposed to any rules and
claimed that there were no laws of music whatsoever. 18 In sum, rules exist for good literature,
art and music but not for beautiful literature, art and music. Beauty evidently adds something
inexplicable which cannot generalized.

After the fall of the West-Roman empire, the contrast between patterns and the unique
continued. History writing in the West was dominated by Universal Histories that consisted
of a unique narrative from the Creation of the world to the Last Judgment. 19 In musicology,
Hucbald (c.840-930) found underlying patterns in polyphonic organa that he formalized in
his Musica enchiriadis. 20 But in other disciplines, such as poetics, art theory, and philology,
pattern searching was far less prominent even though it never disappeared. In poetics the
main goal was to bring textual interpretation in accordance with Biblical interpretation. 21 The
medieval study of art is limited to technical handbooks. 22 And in philology, the Alexandrian

12
For more of these and other cross-cultural comparisons between Europe, China, India and Africa, see R. Bod,
A New History of the Humanities, Oxford University Press, 2013. Due to lack of space, the current essay mostly
focuses on the western humanities.
13
Aristotle, Poetica, XXIV, 60a16.
14
Longinus, Peri Hupsous, I, 3.
15
Pliny, Naturalis historia, 34. 55.
16
Pliny, ibidem, 35.104. Pliny tells the story of the painter Protogenes, who tried to portray a dog foaming at the
mouth. After many failed attempts, Protogenes threw his sponge at the panel in a fury, and this produced
precisely the visual effect he wanted. ‘And so chance gave the painting naturalness!’
17
See Sophie Gibson, Aristoxenus of Tarentum and the Birth of Musicology, Routledge, 2005, p. 169.
18
Flora Levin, Greek Reflections on the Nature of Music, Cambridge University Press, 2009, p. 193.
19
Raoul Mortley, The Idea of Universal History from Hellenistic Philosophy to Early Christian Historiography,
The Edwin Mellen Press, 1996.
20
Dieter Torkewitz, Das älteste Dokument zur Entstehung der abendländischen Mehrstimmigkiet, Beihefte zum
Archiv für Musikwissenschaft, vol. 44, Steiner Verlag, 1999.
21
Alex Preminger, O.B. Hardison and Kevin Kerrane (eds), Classical and Medieval Literary Criticism:
Translations and Interpretations, F. Ungar Publishing Co., 1974.
22
See Erhard Brepohl, Theophilus Presbyter und das mittelalterliche Kunsthandwerk, Böhlau, 1999.

5
method was revived by Lupus de Ferrière (c.805-862). 23 Pattern searching flourished outside
Europe: for instance, in China, Chen Kui (1128-1203) produced one of the most original
works in poetics and rhetoric of all time in which he derived the underlying stylistic rules of
the Confucian classics. 24 His work was widely used as a preparation for the state
examinations during the Song period. In the Arabic world, precise historical transmission
chains, known as isnad, were developed which allowed historians like al-Masudi (896-956)
to write some of the most accurate historical accounts in the middle ages. 25 And in Indian art
theory, the underlying rules for representing foreshortening were formalized. 26

With the advent of humanism, we see a renewed interest in pattern searching in

Europe. In philology, Angelo Poliziano (1454-1494) goes beyond the Alexandrian
philological approach when he takes into account the genealogical relationship between
extant copies. 27 Poliziano realized that a group of completely consistent sources could still be
a problem. Assume that a number of sources – A, B, C and D – all agree on one point, and
that B, C and D are entirely dependent on A for their information. 28 Should B, C and D
nevertheless be included as extra evidence of the authenticity of A? According to Poliziano
they should not: if derived sources were mutually consistent, they should be identified and
eliminated. 29 Sources should be ranked genealogically so that their dependence in regard to
an older source becomes clear. One anomalous manuscript can refute dozens of consistent
manuscripts purely on the basis of its position in the genealogical ranking.

Poliziano used his method with exemplary precision. His quest for genealogies of
manuscripts resulted in highly accurate reconstructions of Terence, Virgil, Seneca, Propertius
and Flaccus. But it is mainly after Poliziano that his philological method revealed some of the
most surprising patterns found in the early modern period. This occurred especially in the
work of the philologist and historian Joseph Justus Scaliger (1540-1609). Scaliger aimed at
unifying all ancient histories (Graeco-Roman, Babylonian, Egyptian, Persian and Jewish) so

23
Robert Graipey, Lupus of Ferrieres and the Classics, The Monographic Press, 1967.
24
Andy Kirkpatrick, “China’s First Systematic Account of Rhetoric: An Introduction to Chen Kui’s Wen Ze”,
Rhetorica, 23(2), 2005, pp.103-152.
25
Ehsan Yar-Shater (ed.), The History of al-Ţabarī, State University of New York Press, 1985-2007.
26
Pushpendra Kumar, Bhoja's Samarangana-Sutradhara: Vastushastra, New Bharatiya Book Corporation,
2004.
27
Angelo Poliziano, Miscellanea, in Opera omnia, three volumes, edited by Ida Maïer, Bottega d'Erasmo, 1970-
1971.
28
This example comes (with slight modification) from Anthony Grafton, Defenders of the Text, Harvard
University Press, 1991, p. 56.
29
Poliziano, Miscellanea, I.39.

6
as to create the definitive historical chronology from the earliest era to his own time. 30 In
doing so, Scaliger not only had to compare many different calendar systems but a very large
number of historical sources too. Scaliger therefore reconstructed various historical texts,
among them Manetho’s list of Egyptian dynasties. Using the information from these sources,
particularly about the duration of the different dynasties, Scaliger was able to date the
beginning of the first Egyptian dynasty to 5285 BCE. To his dismay this date was nearly
1,300 years before the generally accepted day of Creation, which according to biblical
chronology had to be around 4000 BCE. However, Scaliger did not draw the ultimate
conclusion from his discovery, which would have meant that either the Bible or his own
method was incorrect. In order to ‘save the phenomena’, Scaliger introduced a new time
pattern – the tempus prolepticon – a time before time. 31 He placed every event that occurred
before the Creation, such as the early Egyptian kings, in this proleptic time. Scaliger’s
solution may come across as artificial, but for a Protestant in around 1600 it was
inconceivable to cast doubt on the Bible. Yet at the same time Scaliger was too consistent to
give up on his philological method just like that.

But Scaliger’s discovery appeared to be a time bomb. Only a couple of generations

later, an increasing number of scholars – from I. Vossius to Spinoza -- realized that the only
possible interpretation of Scaliger’s result was that the earliest Egyptian kings had actually
lived before the Biblical date of the Creation. This meant that the Bible could not be taken
seriously as a historical source. Scaliger’s pattern of world history conflicted with biblical
chronology, and this triggered a chain of biblical criticism that resulted in the early
Enlightenment. 32

Thus in the early modern era, the discovery of a pattern was no longer neutral, it cried
out for an interpretation. This is not to say that patterns were entirely neutral in Antiquity --
Herodotus and Thucydides also had to interpret their findings, but their pattern (of rise, peak
and fall) corroborated their world view rather than challenging it. In the early modern period,
however, the interpretation of the discovered patterns became critical as they were in strong
opposition to the then accepted world view. Pattern-seeking and interpretation now had to go
hand in hand so as to make sense of the surprising patterns found. This counted not only for
30
Anthony Grafton, Joseph Scaliger: A Study in the History of Classical Scholarship, 2 volumes, Oxford
University Press, 1983, 1993.
31
Joseph Justus Scaliger, Thesaurus temporum, Joannem Janssonium, 1658 [1606], p. 278.
32
At various places it has been shown that there is a direct line running from Scaliger via Saumaise and Isaac
Vossius to Spinoza. See e.g. Jonathan Israel, Radical Enlightenment, Oxford University Press, 2002, and
especially Eric Jorink, Reading the Book of Nature in the Dutch Golden Age, 1575-1715, Brill, 2010.

7
Scaliger’s discovery, but also for discoveries in other early modern humanities disciplines:
from comparative linguistics (e.g. De Laet’s fine-grained analysis that there was no
relationship whatsoever between American-Indian languages and Hebrew, which refuted the
idea that Hebrew was the cradle of all languages 33) to music theory (e.g. the discovery that no
hard distinction between consonances and dissonances could be found, which rebutted the
centuries-old Pythagorean cosmic harmony 34). The same counted of course for the study of
nature: the newly discovered patterns and laws of planetary movements, of falling bodies, of
air pressure and of tiny ‘dierkens’ seemed to be in opposition to the accepted world view and
dearly needed interpretations – and interpretations they got! 35 Thus during the early modern
period, discoveries from ‘humanities’ and ‘science’ went hand in hand in moulding the
sweeping transformation which we now call the scientific revolution. It is hard to find any
conceptual distinction between ‘humanities’ and ‘science’ during this period: also ‘new
scientists’ like Pascal, Kepler, Snellius and Newton studied both nature and texts.36

The first conceptual distinction between a science of the human and a science of the
natural was put forward in Giambattista Vico’s Scienza Nuova (1725) but his work was
ignored for almost a century. It was in the 19th-century that Wilhelm Dilthey (1833-1911)
gave a foundation for the disciplines that we nowadays call humanities. According to Dilthey
the humanities (Geisteswissenschaften) are concerned primarily with verstehen
(understanding) whereas science (Naturwissenschaften) is about erklären (explaining). 37
Humanities scholars would be failing if they observed, counted, measured or hunted for
apparent regularities. What they should be doing is searching for the motives and intentions
of important historical figures. Laying bare these inner mainsprings is more important than
studying the external manifestations of the human mind. In this context one also uses the
distinction introduced by Wilhelm Windelband (1848-1915) between an idiographic
33
Johannes de Laet, Notae ad dissertationem Hugonis Grotii De origine gentium americanarum, et
observationes aliquot ad meliorem indaginem difficillimae illius quaestionis, 1643.
34
See H. Floris Cohen, “Music as Science and as Art: The 16th/17th-Century Destruction of Cosmic Harmony”,
in R. Bod, J. Maat and T. Weststeijn (eds.), The Making of the Humanities, Vol. 1: Early Modern Europe,
Amsterdam University Press, 2010, pp. 59-71.
35
For details, see Eric Jorink, ibidem, 2010; Rens Bod, ibidem, 2013; H. Floris Cohen, How Modern Science
Came into The World: Four Civilizations, One 17th-Century Breakthrough, Amsterdam University Press, 2010;
Penelope Gouk, “The Role of Harmonics in the Scientific Revolution”, in Thomas Christensen (ed.), The
Cambridge History of Western Music Theory, Cambridge University Press, 2002, pp. 223-245.
36
H. Floris Cohen points out that such a distinction may be found back underneath the surface of some of the
work of these ‘new scientists’ – see H. Floris Cohen, “The Natural Sciences and the Humanities in the
Seventeenth Century: Not Separate Yet Unequal?”, in R. Bod, J. Maat and T. Weststeijn (eds.), The Making of
the Humanities, Vol. III: The Modern Humanities, Amsterdam University Press, 2014, pp. 43-52.
37
Wilhelm Dilthey, Einleitung in die Geisteswissenschaften: Versuch einer Grundlegung für das Studium der
Gesellschaft und der Geschichte, 1883, pp. 29ff. For an English translation, see Wilhelm Dilthey, Selected
Works, Volume I, translated and edited by Rudolf Makkreel and Frithjof Rodi, Princeton University Press, 1991.

8
approach to knowledge (which is the study of the unique, the special) and a nomothetic way
of studying (which seeks to generalize). 38 This vision 39 turned out to be extremely influential
as it gave the humanities a powerful identity enabling them to differentiate and emancipate
themselves from the other disciplines. 40

This constitutive dichotomy between humanities and sciences did however not
correspond to actual practice in the humanities. In all humanities disciplines the search for
patterns simply continued, next to the interpretation of these patterns. Also when Dilthey’s
vision was gaining ground -- from the early twentieth century onwards -- there continued to
be both nomothetic and idiographic, both pattern-seeking and interpretative practices in the
humanities. Pattern-seeking components are found not only in linguistics (e.g. De Saussure,
Jacobson) but also in philology (Lachmann, Greg), musicology (Schenker, Lerdahl), literary
theory (Propp, Todorov), art history (Wölfflin, Panofsky), and historiography (Ranke, the
Annales school), just to name a few. This pattern-seeking tradition continued during the
entire 20th century up to the current day.

For example, in art history, the analysis of stylistic patterns was initiated by Giovanni
Morelli (1816-1891) who created detailed taxonomies of pictorial representations of ears,
noses, hands and other parts of the body, as well as clouds, leaves, folds and individual
brushstrokes in Italian art. 41 Whereas Morelli’s stylistic analysis was entirely based on
details, it is thanks to pioneering work by Heinrich Wölfflin (1864-1945) that we have an
analytical stylistic method in which it was not only all the separate parts of the work that
were examined, but also their relationship with the whole, as well as the use of light and
colour. In his Kunstgeschichtliche Grundbegriffe (1915), 42 Wölfflin introduced a gamut of
new stylistic concepts that he grouped in five pairs of opposites in order to characterize style
transitions (in particular from Renaissance to baroque). He defined notions like linear versus
painterly representations, flat versus deep composition, closed versus open forms, clear
versus diffuse representations, among others. He used these notions not only to find patterns
across paintings within a single style but also across styles. His notions still form the

38
Wilhelm Windelband, Geschichte und Naturwissenschaft, 3rd edition, Heitz, 1904.
39
The discussions by Dilthey and Windelband are more subtle than summarized here.
40
See e.g. Gadamer, ibidem, p 6, pp. 56ff. See also the anthology of (abridged) texts in the philosophy of the
humanities, in Jörg-Dieter Gauger and Günther Rüther (eds.), Warum die Geisteswissenschaften Zukunft haben!,
Herder, 2007.
41
Giovanni Morelli, Kunstkritische Studien über italienische Malerei, 3 volumes, Brockhaus, 1890-93.
42
For an English translation, see Heinrich Wölfflin, Principles of Art History: The Problem of the Development
of Style in Later Art, translated by M D Hottinger, Dover Publications, 1932.

9
vocabulary of art historians today, and define what we might call a ‘visual philology’. Yet,
Wölfflin was also criticized by people like Walter Benjamin who in his essay Strenge
Kunstwissenschaft (1933) argued that Wölfflin neglected the social and cultural
interpretations of paintings. 43

Later work by Aby Warburg and Erwin Panofsky did take such interpretations into
account. Panofsky (1892-1968) designed a full-fledged framework for interpretation of art in
Studies in Iconology (1939). His approach incorporates Wölfflin’s stylistic method. Panofsky
defined three levels of analyzing the meaning of a work of art: (1) Primary or Natural
Subject Matter, which corresponded to Wölfflin’s pattern-based approach to form analysis.
(2) Iconographic analysis, which included determining the subject of the painting in terms of
the figures, stories and allegories, and which is based on patterns between pictorial and
literary sources. (3) Iconological interpretation, which contained the deeper significance of
the painting; this called for a ‘synthetic intuition’ that springs from psychological insight and
a thorough knowledge of relevant world views from the period when the work of art was
created. This last level deals with the interpretation of the patterns found in the preceding two
levels. In Panofsky’s view, pattern searching in works of art is thus consonant with
interpretation, where the latter builds on and is a continuation of the former. 44

A similar cumulative construction of patterns and interpretations is found in literary

studies. While it may not be surprising to find pattern-seeking practices in the early formalist
and structuralist literary scholars like Propp, Jakobson and Todorov, it is less well-known that
also those who reacted to structuralism – the post-structuralists – were relying on patterns.
This becomes particularly clear if we look at the work of Roland Barthes (1915-1980), who
built on but also went beyond the long tradition set out by formalists and structuralists. In his
book S/Z (1970) Barthes starts his famous analysis of Balzac’s story Sarrasine with a
structuring of the novella into a complex pattern consisting of 561 reading units (‘lexies’).45
He then analyzes these units in terms of different meaning attributions, showing that Balzac’s

43
For an English translation, see Walter Benjamin, “Rigorous Study of Art”, translated by Thomas Levin,
October, 47, 1988, pp. 84-90.
44
Arrived at this point one could maintain that the identification of a pattern is in itself already a form of
interpretation, and similarly, that an interpretation is a pattern (cf. Michael Baxandall, Patterns of Intention: On
the Historical Explanation of Pictures, Yale University Press, 1985). Yet we believe that for conceptual clarity
it is useful to distinguish between pattern and interpretation, where the identification of a pattern refers to the
first stage of analysis (which in semiology is referred to as syntactic analysis describing the relations between
parts and wholes) and where interpretation refers to second stage of analysis (also known as semantic analysis
corresponding to a meaning attribution to the pattern found).
45
For an English translation, see Roland Barthes, S/Z: An Essay, Translated by Richard Miller, with a preface by
Richard Howard, Farrar, Straus and Giroux, 1975.

10
realistic text is full of symbolic and other connotations which can be interpreted in various
different ways by the reader.

Thus, the meaning attribution to a literary production does not stand in opposition to
but is consonant with the patterning of the text. This is not to say that the opposition between
pattern-seeking and interpretation disappeared from all humanities disciplines. In
historiography, for example, the opposition was strongly felt. While social-economic
historians and (pre-war) cultural historians like Spengler and Toynbee searched for general
patterns in history, their approaches were criticized by narrativists (who argued that only the
‘narrative’ could give an account of an absent past), the critical school (who claimed that only
general criticism could demythologize the past) and postmodernists (who went farthest by
arguing that any claim to historical truth is subject to deconstruction). Yet at a closer look it
appears that the pattern-rejecting historians criticized not so much patterns per se but
universal patterns that were claimed to be culture independent. Their refutation made way for
a quest for different patterns that were culture specific or ideological. And these patterns
needed to be interpreted too. In fact, many historians have found and interpreted patterns in a
historical epoch by employing categories and principles from that period. If a historian knows
the rules of fifteenth-century art theory or rhetoric, for example, s/he can use them to analyze
and interpret works of art, texts and other, even less obvious objects, dating from that time. 46

Also in musicology, linguistics as well as in the more recent disciplines of theatre

studies, film studies, television studies and media studies do we find practices of analyzing,
relating, connecting and comparing humanistic material. And also here, pattern searching and
subsequent interpretation are the order of the day. In film studies, for example, scholars have
developed precise methods of analysing a film by integrating insights from semiology,
literary studies and linguistics. We see this perhaps most clearly in the work of Christian
Metz (1931-1993) who developed his ‘Grande Syntagmatique’ in which he called the
building blocks of film ‘syntagmas’. He provided a hierarchical organization for them so that
the cinematic structure could be visualized and interpreted. 47 Such a cinematic narrative
structure is represented as a tree diagram where the leaves of the tree represent film scenes
and the branched structure reflects the relationships between the scenes, resulting in
syntagmas. This formal analysis into building blocks has led to some surprising patterns. For

46
Baxandall, ibidem, 1985. See also Caroline van Eck, Classical Rhetoric and the Visual Arts in Early Modern
Europe, Cambridge University Press, 2007.
47
Michel Colin, “The Grande Syntagmatique Revisited”, in Warren Buckland (ed.), The Film Spectator: From
Sign to Mind, Amsterdam University Press, 1995, pp. 45-86.

11
example, the narrative structure of the popular series CSI: Crime Scene Investigation, which
has dragged on for years, has been found to consist of only eight narrative building blocks
that are endlessly reshuffled. 48

And in television studies, Raymond Williams identified a widely discussed pattern in

the development of the medium. 49 In the course of the 1970s, TV programmes were no longer
made with separate, successive blocks like news, quiz and film; instead everything flowed
virtually seamlessly into everything else. The natural breaks were now commercials and
announcements about films and quizzes on the following day. This pattern, which Williams
called flow, resulted in a non-stop stream of information, advertising, entertainment and
trailers, which was interpreted by Williams as a way to keep the viewer tuned to a particular
channel.

The most recent branch in the tree of humanities disciplines -- the new and upcoming
field of digital humanities -- has even declared pattern searching in art, music and literature as
its main business. 50 But as has emerged from this essay, the digital humanities simply stand
in a long tradition of pattern searching. The main difference with the traditional humanities is
the digital search in very large amounts of data by means of algorithms. Such algorithms may
on the one hand facilitate the identification of new patterns, while on the other hand they may
neglect patterns that are only observable by humans. Yet the discovery of patterns is in no
way the final goal of digital humanities – these digitally identified patterns need to be
interpreted and criticized too. 51

In sum, if we want to describe the way of working of the modern humanities, it may
be characterized by an integration of empirical pattern searching and successive interpretation
– which is common practice in the sciences and social sciences too. 52 Interpretations do not
appear out of a clear blue sky but are carefully built upon empirical analysis of patterns in
texts, paintings, artworks, musical pieces, films and other human products.

48
Benedikt Löwe, Eric Pacuit and Sanchit Saraf, “Identifying the Structure of a Narrative via an Agent-based
Logic of Preferences and Beliefs: Formalizations of Episodes from CSI: Crime Scene InvestigationTM”, in
Michael Duvigneau and Daniel Moldt (eds.), Proceedings of the Fifth International Workshop on Modelling of
Objects, Components and Agents, MOCA'09, 2009, pp. 45-63.
49
Raymond Williams, Television: Technology and Cultural Form, Collins, 1974.
50
See e.g. Franco Moretti, Graphs, Maps, Trees: Abstract Models for Literary History, Verso, 2007. See also
Melissa Terras, Julianna Nyhan and E. Vanhoutte, (eds), Defining Digital Humanities, Ashgate, 2013.
51
For further discussion, see Rens Bod, “Who’s Afraid of Patterns? The Particular versus the Universal and the
Meaning of Humanities 3.0”, BMGN – Low Countries Historical Review, 128(4), 2013, pp. 171-180.
52
The so-called Discussion sections in science papers can be seen as the interpretation-part of science
publications (Mark A. McPeek, p.c.).

12
 What can we conclude from this longue durée of empirical pattern-searching in the
humanities? It would be mistaken to refer to this practice as a ‘scientific’ way of approaching
humanistic material. 53 Common wisdom has it that the humanities are applying ‘scientific
methods’ when they use a pattern-oriented approach. Instead, we have seen that empirical
pattern-seeking practices have always been part of the humanities, just as they are part of the
sciences. They originated in ancient philology, were revived in the age of humanism and
form the backbone of any well-informed interpretation in the modern humanities. Patterns
and their interpretations are part of all knowledge-making disciplines: while the objects of
these disciplines may differ, their way of working is shared by a profound and age-old
empiricism.

53
For an example of this misunderstanding, see Steven Pinker’s paper in the New Republic of August 2013,
Science Is Not Your Enemy -- An impassioned plea to neglected novelists, embattled professors, and tenure-less
historians, and the many reactions to it. <http://www.newrepublic.com/article/114127/science-not-enemy-
humanities>

13