Narrative Structures in Cross-Linguistic Perspective: English, Hobongan, and Daqan

Four main informational elements have been suggested and studied as central aspects of narrative discourse: causality, character, location, time. The research that scholars have previously undertaken on these aspects has been primarily on IndoEuropean languages, and more specifically on the European side of that language family. The linguistic limitations have indicated that character is the aspect of narrative that readers/listeners attend to most closely. However, in examining narrative discourses from non-Indo-European languages, challenges to the presumed primacy of character emerge. In a partial report on field work conducted in Borneo in 2012-2015, I compare and contrast patterns in the rankings of the four main aspects of narrative in three languages, English, Hobongan and Daqan. I also note the strategies by which the languages make their respective rankings clear, including focus particles (Hobongan), specificity of description (each), and amount of information provided about the aspects (each). I suggest that analyses of the patterns and rankings of information in narrative be included in typological categorizations and linguistic descriptions of languages.


Introduction
Psycholinguistic studies on the ways in which people understand and process narrative discourse have identified four main informational elements of narrative discourse: causality, character, location, and time (e.g., León 2016;Hatzipanayioti et al. 2016;Wassenburg et al. 2015;Mano et al. 2009;Zwaan 1999;Perkins 2009;Zwaan and van Oostendorp 1993;Blanc and Tapiero 2001). The elements must be understood broadly: characters are not necessarily human or animate; temporal information can include aspect, tense, duration, or sequence, among others; causality can be necessary, sufficient, contributory, probabilistic, efficient, formal, final, among others 1 ; and spatial information can be locational, navigational, positional, internal, external, among others. The goal with keeping the definitions of elements flexible is to provide "a more accurate characterization of the linguistic system" (Comrie 1985: 19), rather than to focus on the specific characteristics of any one language.
The aforementioned psycholinguistic studies indicate that readers or listeners do not make equal use of the various elements of narrative discourse. In both English and French, subjects attended most closely to information about characters, particularly as that information related to narrative flow, or causal elements, of the story. Subjects tended to attend less to temporal and spatial information, in some cases recruiting those aspects of information only when those aspects are needed to resolve questions about other aspects of narrative. Subjects rarely attended to spatial information unless directed to do so (e.g., Blanc and Tapiero 2001).
However, my own work (Perkins 2009) suggests that spatial information is important and can be used by authors and storytellers who are skilled in the creation of narrative discourses. Additionally, Graesser et al. (1997) have suggested that a lack of coherence and consistency in spatial information in textoids (short narratives created specifically for use in psycholinguistic experiments) might have contributed to subjects' difficulties in working with spatial information. Further, my field investigations among speakers of several Austronesian languages on the island of Borneo and reviews of language descriptions indicate that the focus on character and the relative lack of focus on spatial information is language specific, with nearly all of the logically possible rankings of the four types of information in discourse being available in the world's languages (Perkins 2017a).
Because few psycholinguistic studies with speakers of non-Indo-European languages exist, it is helpful to use the studies that are available as a starting point for identifying possible strategies for managing elements of information in narrative discourse while being aware that those strategies are probably not universal, keeping open the possibility that additional strategies could be identified as more data become available in a wider range of languages. A variety of cues have already been identified. One type of cue is the level of specificity provided. In English narratives, the most specific information is devoted to character. Another type of cue is the level of coherence and consistency provided: in English, ambiguous pronominal reference with regard to character is humorous at best and more often unacceptable in various ways. By contrast, a lack of coherence and cohesion with regard to spatial information is not readily noticed, which is why textoids can be used in empirical research. A third type of evidence arises from prescriptive 'rules' about creative narratives that are culturally presumed to be preferable to others. One of the main such 'rules' is that authors and storytellers should create well-developed characters (Simons 1996: p.c.). These patterns provide directions to take when beginning analyses of non-Indo-European languages, but they cannot be taken to be an exhaustive list of possibilities.

Assumptions for theory
For this study, I use a linguistic approach proposed as a data-driven way of working with narratives: Nexus Theory (Perkins 2017b). Nexus Theory (NT) is primarily a tool for working with narrative analytically. NT attempts to avoid committing the linguistic equivalent of the Intentional Fallacy (Wimsatt and Beardsley 1946), in which assumptions about shared goals (cooperation, relevance, rationality-making, face management, etc.) are central to the theories used to analyze language. Instead, goals are not assumed, but information about goals is taken to raise open questions to be answered, along with questions about the ways in which elements of narrative discourses such as those included in this study are linguistically instantiated. NT also relies on analyses of components of and interactions between lexical semantics, sentential syntax, discourse structures, and pragmatics. These analyses are data-driven with the goal of investigating and describing these linguistic expressions of conceptualized information in a way that tracks as closely as possible to the patterns that occur within and across languages.
Following Fludernik (1996), I also assume that narrative is in some ways a quality of the presentation of series of events that are cognitively constructed. This is an important assumption for working with narratives that might not appear to fit the usual demands of a narrative, as might happen with non-Indo-European narratives being examined by speakers of Indo-European languages. Assuming that the requirements of a narrative are cognitive and cultural constructs opens the way to a cross-linguistic comparison and contrast of patterns in ways that are relevant within and across languages and cultures. If people in a given culture present and understand a certain set of linguistic structures as a narrative, then that set of linguistic structures is a narrative, even if it does not meet the requirements of a narrative based on some external theory or definition of narrative.
With regard to orality and literacy, it is assumed that written forms represent at least possibly spoken forms. The narratives included in this study were all spoken initially and then written. The narratives have been standardized to the written forms of the languages that currently exist, not including all of the details, such as annotating pauses and vocalized disfluencies. Such pauses and disfluencies could contribute to the performance aspects of the oral narratives, for example, but because this study focuses on the ways in which the four main elements of narrative discourses are managed in language-specific ways, the performance aspects of the narratives are backgrounded and left to future research.
In this study, narrative is defined as a presentation of a story, a sequence of events (Culler 2002: 189). Narrative discourse is taken to be a narrative if it contains each of those four elements: causality, character, space, and time, broadly defined, and if the speakers and hearers of a language present and accept a discourse as a narrative. No commitment is made to fictivity or factivity, and in this study the texts are all factive in order to facilitate relevant comparisons and contrasts. The study is qualitative in order to address fundamental questions that need to be answered before decisions can be made about what can be quantified.
It is assumed in this study that causality is probably the element that is ranked highest in any narrative. Causality provides global coherence (Graesser et al. 1997), and Martin and Schafer (2014) and Lyutikova and Tatevosov (2014) have argued that causality is necessary to the cognitive reasoning involved in the creation of grammatical structures. These studies have been conducted on Indo-European languages, so the possibility of a language that does not primarily prioritize causality remains. The other elements of narrative link more or less closely with causality, depending on language-specific requirements. Identifying and describing some of those links in some languages are the main goals for this project.

Methods and narratives
The type of fieldwork conducted is known as Community-Based Language Research. Czaykowska-Higgins (2009) has described this approach to fieldwork as research conducted on a language or languages, for the language community, with the language community, and by the language community. In other words, the linguist(s) involved are active participants rather than external observers (Dimmendaal 2001), and native speakers are involved in the process of data collection and analysis.
The English and Hobongan discourses were collected during personal interactions. The Daqan narrative was facilitated by a missionary who worked with the Daqan. The people who provided the narratives did so voluntarily and provided permission for me to use the narratives for linguistic analysis. Written consent was not sought because the Austronesian languages involved are primarily oral, which makes written documents and signatures culturally and linguistically suspicious.
The narratives were selected based in part on what was available and on commonalities so that they could be relevantly compared. Each of the narratives is a factive, personal narrative told by a person who was involved in the narrated events. The Daqan and English narratives share a topic: a health event that happened primarily to someone else but with whom the narrator was closely involved as an immediate family member. Each of the narratives was told shortly after the narrated events occurred, and each of the narratives was collected within a single year. Each of the narratives contains culture-and language-specific elements. Three women told the narratives, one speaker per narrative, and those women were all middle-aged or older at the times of the events and the narrations. The narratives are of differing lengths, and the Hobongan narrative does not involve a health event. The narratives do not correlate perfectly, and there are not at this time as many of them as might be preferred, but this is a preliminary study with correspondingly preliminary findings. In addition, the goal of the project is not to say that the phenomena identified in these narratives are always the pattern in these languages, but that, given that these narratives are elicited data but otherwise spontaneous narratives, scholars need to consider more possibilities when analyzing narratives from minority languages, rather than assuming that character is central or primary in most narratives or in most languages.
The narratives have been tidied in order to focus on narrative elements and facilitate readability. As noted, vocalized pauses, background sounds, or other interruptions might be useful for various kinds of analyses, but relative to the analysis of these narratives as narratives, an easier-to-read format is appropriate. Names have been changed in the narratives to protect individuals' privacy. The interlinear versions of the narratives have been split up by sentences to make clear the relationships between morphemes and meanings but have not been numbered as individual sentences in order to facilitate thinking of the narratives as larger-than-sentence units of language (Pike 1964).

Details on Narratives
I have used a number of abbreviations in the non-English narratives: FOC = focal element marker, DEP = dependent clause marker, inc = inclusive, tri = trial, LOC = location marker (not locative because Hobongan and Daqan both lack case marking), EMPH = emphatic marker.
The Hobongan narrative would not be a narrative by all definitions of narrative. For example, Fludernik's emphasis on ′experientiality′ would render this discourse a report rather than a narrative. However, it must also be kept in mind that, for the purposes of this study, I have defined ′narrative′ linguistically according to information that is contained within discourses, rather than relying on an ′experience′ that might be external to a written or spoken discourse. In addition, these are the written form of the narratives as told and accepted by native speakers. Originally, the narratives were spoken, and there were complications and audience-participations that I have omitted in order to keep the written forms readable. By the definition used here, these are narratives in their written forms, and by many definitions, in their spoken forms, as well.

An English narrative
I wanted to sleep in this morning, but Lois woke me up at 6:30 because she'd cut herself. Her parents gave her a pocketknife for her birthday, and when she woke up early and couldn't go back to sleep, she went down in the basement to work on one of her craft projects. And of course she cut herself. So she came up to get me, and she's so bad at blood. She said she was sorry about waking me up, but she thought she was going to throw up, and wanted to go sit by the toilet. I told her to sit down right there and the first thing we'd do was put a band-aid on the cut. I went to get the bandaids, and she sat there, and I got the band-aid on her so she couldn't see the blood. Then I had her lie down, and of course she felt much better in seconds. She's so bad at seeing blood. It's kind of funny.
Causality in this narrative is primary and complex. The layers in the causality interact through the characters and their interactions with one another. The narrative begins with a causal explanation of why the narrator did not sleep later into the morning. The causality extends into why the Lois character had awakened the narrator. The narrator inserted an additional layer of causality by explaining why Lois had cut herself, both by noting the means and the opportunity for the cutting. After two sentences in the narrative, there are at least three layers of causality, all tied to characters in the narrative, and at least two specific forms of causality, means and opportunity. The layers of causality introduced in the first two sentences of the narrative are anaphoric, explanations provided after the elements that they explain.
Character in this narrative is perhaps most important to the flow of causality in the narrative. The narrator and Lois are described with the most detailed and specific information. Lois is established as a creative and independent person, and the narrator establishes herself as having at least the medical competence necessary to perform first aid and to recognize interpersonal and intrapersonal dynamics of a situation, while evaluating the situation from a meta-perspective that is facilitated by the after-the-fact narration.
Temporal information occurs with relatively specific references: 'this morning,' and '6:30.' After that initial introduction, temporal references refer anaphorically to the more specific temporal elements given in the introduction and are developed sequentially only via pragmatic implicatures that not everything can happen simultaneously. The simple past tense used for the events in the narrative places all of the possible temporal details in the same grammatical category, and real-world knowledge about sequence, as well as some of the sequences of the narrative, are used to make distinctions within the simple past. The sequence of the narrative is interrupted twice, both times for the narrator to make an observation about Lois. Both interruptions are signaled grammatically by the use of the habitual present tense, which is used nowhere else in the narrative. No duration is given for this narrative, except as an estimated subdivision of another event, with the generic 'seconds.' Spatial information is the least developed in the narrative, but it is included in the narrative, and it is important. The most specific location is the basement, which is noted to be 'down' from the narrator's pragmatically inferred location, a location that is not named or described in the narrative. Beyond the basement, a generic 'there' provides Lois' location, otherwise unspecified. There is a 'toilet', but no description of location is provided beyond real-world knowledge of where such fixtures are usually located. Several scales are included: the personal scale of posture, the just-beyond-personal scale of locations, such as the inferred positions of 'there' and 'toilet,' and a broader scale of an inferred house that contains a basement, the non-basement parts of the house, and a room at least for the toilet. No geographical-scale information is included. Navigationtype information is not detailed but is mentioned with Lois' movements down to the basement and back to awaken the narrator. The navigation is linked closely with the character's actions and can be inferred to take place with regard to the given or implied locations, but the relationships between and among locations are not specified.
Overall, this narrative is a solid example of typical factive English narratives. Causality and character are closely linked and well-developed with detailed and specific information that is not ambiguous within the narrative. Time is introduced specifically but then relies on real-world knowledge for relevant sequencing. Duration is mentioned but not with a specific description. Spatial description is generic and relies on real-world knowledge of how most houses are constructed. Had the narrative been provided outside of an Englishspeaking, middle-class context, the spatial information would have been difficult to parse. Ambiguous references to 'there' are not disambiguated or described later in the narrative, directly in contrast with the ways in which causality was developed. Causal information is primary in this narrative, but less explicit and more implicitly bound to the locations and navigational processes given in the narrative. The travel up the Hobongan River provides the causal and navigational outline of the rest of the narrative. This kind of 'abstract' at the beginnings of narratives in oral cultures is relatively common but is not always required. The causality interacts closely with the locational and navigational information. The fact that some rapids were circumnavigated on foot provides inferable information about the severity of the rapids: the rapids were circumnavigated on foot because the boats had to be mostly emptied in order to be cantilevered over and around some of the rocks. Listeners (readers) who are familiar with the river would understand this kind of inferencing pattern. A parallel instance for locations is the grave site. The place was where the coffins where, but there was also a large ceramic jar. The causality does not need to be explicitly stated because those who are familiar with Hobongan burial sites would know that people's possessions were often disposed of along with the bodies, so it is not surprising to see other objects. The notable factor with the jar was that it was especially large, which is why the size is marked with an emphatic.

A Hobongan narrative
Locations and navigational processes are complex, layered, and given specific detail and names, in some cases. The first gravesite visited is described geologically (the cliff), geographically (its place along the route described in the narrative), and with the proper name that the Hobongan assign to this particular grave site: Boboq. 'Upriver' is used both as a location and as a direction of travel. The woman who told the narrative assumed, and allowed the narrative to afford the inference, that the starting point for the day's travels was not considered to be 'upriver'. This inference can be made based on the content of this narrative, although it is also possible that those familiar with the Hobongan world would know that the current villages are downriver from the original Hobongan territory. The teller made reference to a fact of Hobongan locational history in using 'upriver' as a location, not just as a direction of travel. Locations can also be subdivided into smaller locations for more specific description, providing a variety of scales. The teller mentioned the big rapids as a single location, but she subdivided the upper part of the big rapids to note that the upper part was where the boats had to be maneuvered around and over rocks.
Mang Lo is also a location with cultural and agricultural importance. It is a rock that traditionally was used as part of a landscape-scale orientation to determine what time of year to plant rice. Relatively speaking, it is a trivial rock in the river when compared with some of the rocks that made the rapids that had to be walked around. Its cultural importance in the larger landscape provides it with salience. The gravel bar where people ate the main meal of the day is a much larger feature in the river but is named with reference to the smaller, more culturally salient rock.
In contrast with the detail and specificity provided for the locations, the pronominal references in this narrative are sometimes difficult to track within the narrative. The woman who told the narrative most frequently used first-personinclusive pronouns, but there were some shifts that are not always clear, such as the shift between 'we' and 'they' in the fourth sentence. She separated some of the group from the group that she was including herself in without specifying on what basis the split was made or who was included in each of the groups. It could plausibly be suggested that because this narrative was told about an event that the people listening to the event were familiar with, further specification was not necessary. That is the case for the inclusive first-person material, but the reasons for the split are not clear to me as a person who was there and knows the people involved, and it is not syntactically clear in the narrative itself. However, in the section about the big rapids, the subdivision is clear to those who are familiar with the events described in the narrative, and this helps to clarify the split in the pronouns. When the Hobongan encounter unnavigable rapids, the women and children walk around the rapids, and the men maneuver the boats.
This kind of Hobongan-real-world-implicature is available in the narrative, but not always. When such implicatures are available, they cooccur with a highly specified subdivision in the locational information in the narrative. However, making assumptions for people who already know is not a strategy that is used for the locations: both Boboq and Mang Lo, although intimately familiar to the people who heard this narrative and who were involved in the outing, are named specifically and given descriptive details that do not occur for any of the people involved. The most detail that is given with regard to the people (characters) in the story is when a couple of people, who would have been clear to those who were involved, are separated out as walkers. The teller named only one of them (Mari), but she addressed two of them in the narrative, even though other people were present at the telling. Given the scenario that the teller described, it is unclear why she separated out or addressed the two people she did. Those people were not the only ones who walked around those rapids, and the two people who did walk around the rapids were in different boats from the teller, and therefore in a different subset of the group. The teller did not separate herself from the people in her subset, but from a different subset. No explanation is given for this in the narrative, and the situation, even though familiar to me, does not clarify this choice. For those who like statistical information, locations are named three times more often than people.
Temporal information is linear in this narrative but not always accurate to what happened, which would need to be sorted out by the people listening to the narrative. The teller did not provide any hint that she was rearranging events or any explanation for rearranging the events. For example, people prayed after making the rice but before eating at Mang Lo, rather than after the meal and before traveling, and the rice-cooking and shelter building have been linearized here in ways that are not entirely accurate descriptions of what happened. The teller might have forgotten these details in their sequence, but it should be noted that all of the locations are given in precise geographical order as they were encountered along the navigational route. The teller never forgot, inaccurately linearized, or randomly inserted what was crucial to the narrative but only the lowest ranked information in Hobongan narrative discourse.
The Hobongan narrative differs significantly from what would be expected of an English narrative, providing detail and resolution on locations and navigations, rather than about people. The familiarity of the listeners with the situation did not appear to affect the naming of people or places. Instead, places and people were given detail and names according to patterns in Hobongan narrative, not according to what would be most informative to the people listening to the telling. Temporal information is treated almost as an afterthought, which is how it can be non-linear despite the linearity of the spatial information.

A Daqan narrative 2
Waktu aku Botohe Said umoro 5 peramaq io In this Daqan narrative, causal information links most closely with temporal information, specifically duration and sequence. The duration of the pregnancy frames the narrative, and some elements of the narrative that seem odd if assuming a character-focused narrative make sense when the duration of the pregnancy is the focus and the reason for the inclusion of those elements, such as having the pig and giving money to relatives who don't otherwise feature in the story.
The sequence of happenings within the duration of the pregnancy follows almost exactly the order of events, even when that order makes some of the pronouns ambiguous, as with 'his' referring to the pig's father or Angus' father. The only real interruption in the sequence occurs at the end of the narrative, when the storyteller interrupts the sequence to comment externally to the narrative, analogous to such comments in both the English and the Hobongan narratives. The comment can occur outside of the sequence of the narrative because the comment is outside of the events of the narrative. The storyteller shifted out of the narrative sequence to suggest that her memory might not be as exact on the duration of her brother's stay as she might prefer. The interruption of the sequence provided a conclusion to a duration (length of stay) that was related to the duration of the pregnancy and provided a reason for the storyteller to end the story where she did: she concluded both of the durations that she had introduced.
A couple of possible interruptions in sequence occur because the sequence of the narrative and the sequence of the clauses is not necessarily parallel. Clauses in language are necessarily linear, but the unidirectional language can be used to describe events that occurred simultaneously or partially simultaneously. The exchanges of money are a good example. The money events occurred in sequence with everything that was happening with regard to the mother's illness, and they are included because of their place in the sequence, not because of their connection to the mother's illness: the sequence of the illness and pregnancy are what explain the contents of this narrative, not the characters as such.
The information about the characters involved in the narrative is not so closely managed or so precisely presented as the sequential information. As noted, the 'his' for the father is one example of a possible ambiguity (syntactically ambiguous), probably resolvable with the pragmatic knowledge that a human father is part of the narrative already (pragmatic resolution of the ambiguity).
That an unborn child, Sadi, can be a character as much as any other character in this Daqan narrative pragmatically suggests the importance of children in Daqan culture.
There could be some question with regard to the open house and the appearance of a long-dead grandmother in the story, but that is part of the traditional belief system remaining in Daqan culture. Sometimes a dead relative will appear, presumably to accompany someone who is about to die.
Spatial information has the least detail and specificity in this narrative. The mother's house is different from the original location, and those are the only major locations in the narrative. People's postures and movements are not specified, and navigation between locations is not mentioned. The locations that are given are not described, but the typical elements of the locations, such as doors for houses, are pragmatically assumed to be available. There is little, if any, link between spatial information and causality, even where it might be expected. The father left to work on the rice harvest, and that is inferable from the narrative to those who are familiar with the ways in which the Daqan harvest rice.
As with the Hobongan narrative, Daqan challenges expectations that would be in place for character-based narratives. Without that assumption of characterprimacy, the narrative can be analyzed on its own terms, in which temporal information, especially duration, is primary. Duration-primacy explains the inclusion of some of information, such as the information about the money and the pig. The narrative ends when the duration ends, not when the main character ends.

Discussion and conclusions
Each of the narratives exemplifies broad cultural preferences for organization and presentation of information in narratives. Although authors at least in English play with these preferences in order to achieve literary effects of various types, it should be kept in mind that the possibility of playing with preferences is additional evidence that those preferences exist in certain ways in certain languages. As literacy increases in Hobongan and Daqan, it would be an interesting area of further research to investigate how writers in those languages come to play with their respective linguistic and cultural patterns. Recently collected preliminary evidence from Hobongan suggests that there is a phase of writing oral literature in which the usual patterns are not yet played with but written down exactly as they occur in spoken narrative.
In all three of the narratives, the storytellers included themselves as the tellers of the narratives in the first clauses of the narratives. In English, the storyteller put the first-person pronoun as the first word of the narrative. In Hobongan, the storyteller situated the narrative in geographical space before noting her role in the narrative. In Daqan, the storyteller began by placing herself within the duration that governed the entire narrative.
In all three of the narratives, the storytellers interrupted the sequences of events to comment on the narrative. The interruptions related to the information in each of the narratives that linked most closely with causality in the narrative. In the English narrative, the interruption allowed the storyteller to develop herself as a character in the narrative, by giving information about her evaluation of the events in the narrative. In the Hobongan narrative, the interruption addressed the audience with a question about spatial information. The Daqan narrative provided information about the narrator's uncertainty about a secondary duration.
The narrative with the fewest characters was the English narrative, with only two (and the implied audience who was addressed in a comment). Those two characters were closely tracked informationally, with no instances of ambiguous pronouns and with some development of the characters, providing emotional and physical description of the characters with regard to their involvement in the events of the narrative. The narrative with the most characters was the Hobongan narrative, with the narrator and a few people who had been present at the events described as well as at the telling of the narrative being the main characters and nearly forty people who were not described in any detail being secondary characters. Between the English and the Hobongan narratives on the character continuum is the Daqan narrative, with a couple of primary characters, and several secondary characters who are named or described but not in detail.
The narrative with the most information about location and narration was the Hobongan narrative, in which the sequence of events was interrupted not only so that the storyteller could comment on the material but so that the location of one of the main characters and his navigational route to the location at which the main conversation took place could be specified. Neither Daqan nor English included much spatial information overtly, leaving much of that information to be inferred from real-world knowledge of the ways in which houses are built or rice is harvested.
The narrative with the most specific and detailed temporal information was the Daqan narrative, in which the main events of the story could be backgrounded temporarily in order to reinforce the sequence within the specified duration. Sequence was followed in the English narrative with the exception of the out-ofnarrative-sequence comment. In the Hobongan narrative, temporal information was backgrounded, leaving the sequence to be inferred in some instances based on the audience's knowledge of the events.
Each narrative, within each language, prioritizes different types of information. Those priorities are indicated by means of the level of specificity of information provided for each of the various kinds of information with more important information being described more specifically and with more detail.
Storytellers also indicate informational priority by constructing coherence and cohesion for some elements of the narrative but not for others. In an English narrative, maintaining coherence and cohesion for character can be accomplished at least in part by making clear who the characters are and minimizing ambiguous pronominal reference. This contrasts with Daqan and Hobongan, in which syntactically ambiguous pronominal references and unspecified groups can occur without damage to the coherence and cohesion of those texts. In a Hobongan narrative, maintaining coherence and cohesion can be accomplished at least in part by making clear what the locations are, and how a main character or characters in the narrative navigated to important locations. In Daqan, temporal coherence and cohesion can be achieved in part by noting a duration during which the events of the narrative occurred and by including a full sequence of events within that narrative, including secondary events.
Future research could include collection and analysis of additional texts from these and other languages. Cross-linguistic narratology is in its infancy, but the potential contributions to a number of fields, including typological linguistics, narratology, and literary studies, and the recognition of the importance of narratives in languages and cultures and the conservation of languages and cultures around the world suggest that such analyses are increasingly important. Analyses of the patterns in which languages manage and prioritize aspects of narrative could be a relevant and informative component of language descriptions, making those descriptions more descriptive and less prescriptive by acknowledging and including significant work on the level and type of language that appears to be more significant to native speakers and cultures than morphology or syntax: narrative.