Sequentiality is the linear, unidirectional succession of elements or events, either reversible (as with motion in space) or irreversible (as in the flow of time). Regarding language texts, it is the irreversible sequence of signs from beginning to end, i.e. the succession of “these words in this order” (Cameron 1962: 145). In narration it is the sequence in which events are presented (discourse); also essential are the sequence of the narrated events in the real or imagined world (story) and the sequence in which they are received by listeners or readers.
Sequentiality is an essential trait of narration and narrative on every level of investigation: on that of the author or production (in terms of the processes of telling or writing); on that of the presentation or discourse (in terms of the sequence of signifiers); on that of the presented events in their individual shape or story (in terms of their chronological or otherwise meaningful sequence); on that of the underlying bare scheme of successive narrative events or fabula (as defined by Bal  1997); on that of the listener or reader (in terms of the processes of listening or reading); and on that of the relations between these levels. Closer inspection reveals that the identity of any narrative depends on its particular sequentiality on all these levels—with the exception of the process of writing on the level of production.
Verbal narration is a temporal medium: telling or writing as well as listening or reading are temporal processes that take place in time, and the real or imagined world in which narrated events are placed is held to be governed by temporality. Consequently, the important relation between the sequence of presentation and that of presented events is mostly discussed in temporal terms. In the case of a concurrence of these sequences, one speaks of ‘chronological’ telling; otherwise, ‘anachrony’ obtains (Genette  1982). This seems logical and is indeed perfectly adequate with regard to oral narration. However, when narratives are recorded or written, they are to a considerable extent freed of that temporal anchoring. Not only can they be made accessible at different times in different places in the story, but it is also possible for listeners or readers to deviate from the notational sequence of presentation and approach the various parts of a narrative text in a sequence of their own determination. Although authors usually structure their narratives in the expectation that readers will read them in the sequence in which they are written, they also know that, for instance, more than a few readers of detective stories read the ending early on in order to know ‘whodunnit.’ What is rarely realized is that, owing to the dynamic character of the construction of meaning in the process of reading, such a change in the sequence of reception actually results in the creation of a new text from the elements of a given one—hence, of a different narrative of the reader’s own making. It follows that it is more appropriate to discuss discursive sequence in terms of sequentiality than in terms of temporality.
On closer inspection, this can also be seen to hold true regarding the order of events on the story level. Of course, the assumption that, in analogy to real-life experience, this order must be temporal—events occurring either simultaneously or in chronological sequence—is well founded, and the construction of this temporal order in the process of listening or reading is not only commonplace but also conducive to the further assumption that succession implies causality (Chatman 1978; Kafalenos 2006; Pier 2008). Temporal sequence is not the only way of meaningfully arranging narrated events, however. Sternberg (1990, 1992) has listed the further options of simultaneity (e.g. in terms of addition or alternation of multiple strands), non-temporal sequentiality (as in case of a hierarchical order, e.g. proceeding from the more general to the particular, or vice versa), functional sequentiality (e.g. in terms of multiperspectivity), and ‘suprasegmentiality’ (e.g. ‘spatiality’ or, rather, a strategy for creating the illusion of spatial arrangement). It has also been shown that with “polychronic narration” (Herman 1998, 2002), the creation of order becomes problematic.
Thanks to the impressive amount of historical writing, from diaries, autobiographies and biographies to chronicles and historiography proper, as well as to fictional autobiographies, biographies and historical tales and novels, the most common kind of narrative sequentiality is concurrence between the sequence of presentation and the temporal sequence of narrated events or ‘chronology’. From Thucydides in the 5th century BC to Gibbon in the 17th and Ranke in the 19th century, the writing of history has been marked by a chronological presentation of the past “as it really was” (Ranke  1909)—at least in intention. The same holds true for biographical writing from late antiquity (Plutarch, Suetonius, Tacitus) to the present day, for fictional biographies and autobiographies such as Graves’s I, Claudius (1934) and a host of fictitious biographies from Grimmelshausen’s Simplicissimus (1669) and Defoe’s Robinson Crusoe (1709) onwards. Even when the presentational sequence in a narrative frame happens not to be chronological, it is normally thus in the framed stories (of which there are many examples from Chaucer’s Canterbury Tales (1378) to Brautigan’s 1970 postmodern Trout Fishing in America).
On the other hand, ever since the Homeric epic (8th century B.C.), authors have deviated from this concurrence between the sequence of presentation and that of the events told; the Homeric method of beginning in medias res was long the norm for the epic genre. A yet more audacious and even playful experimentation with narrative sequence was encouraged by the novel because of its rather loose generic form. Significant interruptions of chronological telling can be found in Britain from Fielding onwards; Sterne’s Tristram Shandy (1767) is already an example of deliberate and excessive play with presentational sequence.
It was, however, only with the turn to modernism, when authors began to write with the deliberate aim of making reading difficult, that experimenting with the sequence of presentation became a prominent feature of literary narrative. A major tendency was to write “against time” (cf. Grabes 1996), a practice that led to weakening sequential structure through fragmentation and collage and the concomitant deviation from the traditional rule of apprehension. Well-known examples of what Frank (1945, 1991) was to call “spatial form” are Belyj’s cubist novel Petersburg (1913) with its attempt to achieve a ‘spatial’ effect, Faulkner’s sequential creation of the effect of synchronicity in The Sound and the Fury (1929) and of multi-perspectivity in As I Lay Dying (1930), and the use of the collage by Dos Passos in his trilogy USA (1930–36) to display a broad panorama of the United States during the first three decades of the 20th century.
Even more radically experimental was some postmodern writers’ willful avoidance of a definite narrative sequence by adopting particular kinds of notation, a strategy exploited by offering the option of reading the parts of a literary text in several sequences and thus creating several works on the basis of one written text (cf. McHale 1987: 190–93,on the “schizoid text”). In the blurb to the 1962 Weidenfeld & Nicholson edition of Nabokov’s Pale Fire, the book is fittingly called a “do-it-yourself detective story” because the reader can choose among the several possible sequences in which the various parts of a novel (disguised as a critical edition of a narrative poem) are read. Cortázar, in the introduction to his novel Rayuela (trans. Hopscotch, 1966), suggests two sequential patterns, making two works created from the same text. A similar strategy is used by Michel Butor in Boomerang (1978) by having parts of the text printed in different colors, so that one may either follow the sequence of parts in one color only or read the whole book without paying any heed to the colors. Even in such cases, however, “[t]he author is the one who proposed a number of possibilities which had already been rationally organized, oriented, and endowed with specifications for proper development”(Eco  1989: 19). This, however, no longer obtains when the potential sequences of apprehension become practically countless, as was the case when experimentation reached a further stage. B.S. Johnson, for instance, went so far as to present the 29 chapters of his novel The Unfortunates (1969) loosely in a box, with only the first and last marked as such, and Mark Saporta in his Composition No. 1 (1962) offered his readers no more than a batch of unnumbered pages, printed on one side, which, according to the author’s instruction on the cover of the box, “may be read in any order. The reader is requested to shuffle them like a deck of cards.”
A much more elegant and widespread option was the appearance of hypertext, an interactive kind of textuality, embedded in a digital environment, which permits the generation of multiple stories by leaving the sequential order of textual elements or events to the ‘reader’ (or, rather, co-author). In a Storyspace hypertext like Michael Joyce’s afternoon: a story (1987), a great number of textual units, called ‘lexias’, were offered for various kinds of arrangement, though conditions could be placed on the activation of links. In such web-based HTML-Frames from the 1990s as Olia Lialina’s My Boyfriend Came Back from the War (1996), a computer page is subdivided into several frames that remain synchronically given, so that the sequence in which they are explored and manipulated is left to the discretion of the user. In artificial intelligence-driven texts like Façade (2002) by Michael Mateas and Andrew Stern, not only can the user act as co-author but he or she can ‘enter’ the created artificial world and move around in it like a character in a play. No wonder that particular kinds of computer games are also regarded as narratives!
What must be kept in mind, however, is the fact that both a particular sequence of presentation and a particular sequence of events are such important structuring devices of a narrative that we are actually dealing with different narratives when one or both of these sequences are changed. Postmodern textual experiments like those by Johnson or Saporta and, even more so, the various kinds of hypertexts and computer games, are therefore programs for the creation of a multiplicity of narratives, not narratives per se.
Attention to sequence was already given in the rhetoric and poetics of classical antiquity (Cicero, Rhetorica ad Herrenium, 55 B.C.; Horace, Ars poetica, 14 B.C.) when dealing with disposition, or the order of speech, and the distinction between ordo naturalis and ordo artificialis, thus anticipating the later differentiation between chronological and non-chronological narrating. Another, albeit indirect, focus on sequence is found later in Lessing’s Laokoon (1766 ), where literature is defined as a temporal art, the domain of which, from a mimetic point of view, is therefore the rendering of actions and events. And although this definition may be too narrow to apply to literature in general, it fits well for narrative, where the presentation of some sort of change seems indispensable.
Beginning with the discovery by Propp (1928 ) that, in a corpus of 100 Russian folktales, invariant functions are linked in particular sequences, sequentiality—often under the name of temporality or causal sequence (in the sense of causa efficiens)—has played a crucial role in the various attempts to explain narrative structure, ranging from the assumption of a “universal plot” (Raglan 1936) or a “monomyth” (Campbell 1949 ) to that of an underlying “narrative grammar” (Todorov 1969; Greimas 1971), and the models inspired by generative-transformational linguistics (Prince 1973; Pavel 1976, 1985; van Dijk 1976; Doležel 1979) and schema theory (Rummelhart 1975; Schank & Abelson 1977; Ryan 1991; Herman 1997; Emmott 1997; Hühn 2008). This pertains in particular to the relation between the chronological sequence of told events and the sequence of their presentation in a narrative, a relation discussed by the Russian formalists in the 1920s in terms of the distinction between fabula and sjužet and defined by Genette (1972 ) as that between histoire (or narrative content) and récit (or narrative discourse). Using different terms for distinctions already made by Müller (1948 ) and Lämmert (1955), Genette ( 1980: 35, 40) distinguished between “the temporal order of succession of the events in the story and the pseudo-temporal order of their arrangement in the narrative,” drawing attention to “anachronies,” or “forms of discordance between the two temporal orders of story and narrative” that result from “analepses” (flashbacks) and “prolepses” (flashforwards), as well as to the special case known as “achrony,” an incident deprived of any temporal anchoring in the chronological sequence (84). Richardson (2002) was able to show how great the variety of temporal sequences of told events in a narrative can actually be: circular, contradictory, antinomic, differential, conflated, and dual or multiple. He also introduced the term “metatemporal” for unusual or impossible temporalities (or, rather, sequences which cannot be imagined in terms of temporal relations). A narrative text allowing the imagining of multiple sequences of events became for Barthes in S/Z (1970 ) the ideal of a “writerly text” that turns the reader from a consumer into a producer. After a few attempts by authors to reach this goal with printed texts by leaving the sequence of discourse undetermined to some extent, readers generally become co-producers with the advent of interactive hypertexts that require subjective choices or interaction between concurrent processes instead of following a particular, fully author-determined line of discourse (cf. Ryan 1991, 2006). Even computer games, with their guided text production yet still wide-open sequentiality, are treated by some narratologists as narratives (→ Neitzel: Computer Games; cf. Ryan 2006).
More recently, the emphases of structuralist narratology have been influenced by including the perspective of the reader, who is held to construct the fabula from the discourse, and this has led to bringing the traditionally more text-oriented structuralist view of sequentiality into line with the theory of reading (see Rimmon-Kenan 1977; Sternberg 1978; Perry 1979; Bremond  1980; Culler 1980; Ryan 1991 and 2006; Ireland 2001; Dannenberg 2004; Kafalenos 2006; Baroni 2007; Pier 2008). Sternberg, Ireland and Kafalenos, in particular, have extensively discussed sequence in a way in which the structuralist paradigm is opened to the more comprehensive one of a theory of reading.
While the linearity of the discourse of written narratives (cf. Miller 1992) suggests reading word by word, line by line, and page by page according to the culturally determined code of writing and reading, it is often forgotten that, once written, all these words, lines and pages of a text exist synchronously in a spatial arrangement and can be approached in a temporal sequence more or less determined by the reader. This ‘spatiality’ of the written text tends to obscure the fact that a particular discursive sequentiality is an essential feature of the identity of a narrative and that readers who do not stick to the conventional rule of approaching the text in the sequence presented by the author will actually read a different narrative of their own making (Grabes 2013). This has to do with the dynamic character of sequential perception: apart from the ‘primacy’ and ‘recency’ effect (Luchins 1957; Perry 1979), cognitive psychology has shown that later deviations from schemata or semantic models, once they have been introduced, tend to be distrusted or wholly overlooked (Hovland, ed. 1957; Grabes 1978). To be reckoned with is what Iser (1976 ) has termed the “wandering viewpoint,” with its continuously changing “retentions” (or looks back at what has already been read) and “protentions” (or assumptions of what may happen in the future; cf. Toolan 2004, 2009) and the ongoing process of re-evaluation motivated by new information (cf. Grabes 1978, 2004, 2013; van Dijk 1980; Sternberg 1992, 2001; Margolin 1999; Kafalenos 2006).The stance taken by earlier structuralist narratology that tends to keep in view all features and parts of a narrative “at the same time” is therefore an artificial one designed to obtain valuable scientific insights but which remains far removed from the actual experience of reading a narrative. The perspective of the reader had already been introduced by Kermode (1967, 1978) and by Eco (1979), but the processual and dynamic character of imaginary worldmaking during reading has also been investigated by reception theory, following up the insights of Ingarden (1937 ), Iser (1976 ), and Ruthrof (1981). Also to be mentioned are the investigations into the creation of fictional “characters” during the reading process (Grabes 1978, 2004; Margolin 1987, 1995). With the advent of interactive hypertexts, the processual character of narration has by necessity become foregrounded, since the ‘reader’ or consumer is also a ‘writer’ or producer (cf. Barthes 1970 ;Yellowlees 2000). This is a field which is still under discussion, and the growth of digital media will ensure than it remains in focus. A most useful survey of the various forms of interactive narrative texts has already been provided by Ryan (2006).
The more recent tendency of having authors present their narratives in the form of public readings has strengthened the general awareness that narration is a processual and thus a sequential art, and it is to be hoped that this aspect of narration (“performance-culture orality”) will receive more attention in narratological research.
One question that deserves more attention is the sequential building up of varieties of imaginary space and of more or less static objects in space. The fact that the particular sequential creation of such varieties in discourse has an impact on the imagined products has not been sufficiently studied. It will thus be fruitful to develop models of how particular sequences of signs are transformed into synchronous systems of meaning by processes in which each succeeding stage incorporates all previous stages. Due to the sequentiality of language in general and its importance for the study of literature (cf. Grabes 2013), such models are of particular interest for narratology, since the sequences of signs in narratives, which can be quite lengthy, require potent strategies for keeping track of earlier stages of the story and revising them as the text advances. They are, in fact, indispensable, not least for the critic, who, in interpreting particular works, must remain highly attentive to sequence—not merely in detective stories but in narratives at large. Among the early attempts to tackle this problem was work on the primacy-recency effect (Luchins 1957; Perry 1979), but also Holloway (1979) who, using set theory, described each phase in the reading process as a set of sets, a continually actualized reconfiguration of the entire sequence of information up to that point. More recently, important aspects of this process have also been studied from a cognitive perspective by Emmott (1997) on the basis of “contextual frames” and from a text semiotic perspective by Pier (2004) in his study of “narrative configurations” as well as within the framework of “narrative progression,” notably by Toolan (2004, 2009) employing advances in corpus linguistics, and by Phelan (2007) in his work on rhetorical narratology.