and generalizing statives
. | Narrative mode . | Argumentative mode . |
---|---|---|
Entities | Events and states | Primarily abstract and generalizing statives |
Temporality | Located in time | Atemporal |
Advancement | Events related to each other in (narrative) time | Metaphorical motion through the domain |
Tense Interpretation | Continuity and limited anaphora | Deictic |
Source; Adapted from Smith (2001 : 197–202).
Smith (2003) posited five common discourse modes in written texts: narrative, descriptive, reporting, informative, and argumentative/commentary. Discourse modes often align with genres. For example, the narrative discourse mode is paramount in the narrative fiction genre and the argumentative–commentary discourse mode is common in the editorial genre; however, a text can employ multiple discourse modes (‘the discourse modes cut across genre lines’; Smith 2001 : 185). For example, the paragraph below transitions from an argumentative to a narrative mode (using an MDM strategy):
[Argument:] I feel reasonably certain of the final verdict on the current impeachment affair because I think history will see it as the climax of a six-year period marred by a troubling and deepening failure of the Republican party to play within the established constitutional rules. [Narrative:] It was on Election Night 1992, not very far into the evening, that the Senate minority leader, Bob Dole, hinted at the way his party planned to conduct itself in the months ahead: it would filibuster any significant legislation the new Democratic President proposed, forcing him to obtain 60 votes for Senate passage ( Ehrenhalt 1998 , cited in Smith 2001 : 203, notation added at the beginning of the sentences).
The first sentence employs an argumentative discourse mode with a first-person proposition of belief (‘I feel …’), supported by reasoning using another belief (‘because I think …’). Then, it shifts to a narrative discourse mode in the second sentence; this is marked by temporality and events related by verb tense and aspects (‘It was …’) that are common in story narrative genres. Smith (2001) argues that transitions between discourse modes are universal across languages.
A social practice approach to language learning ( Barton 2007 ) might explain the origin of students’ MDM writing strategies. Language learning often occurs when people use language in their daily lives rather than in a vacuum or in the order of grammatical forms taught at school ( Barton and Lee 2013 ; Barton and Potts 2013 ). In addition to classroom instructions in school, students constantly engage in meaning-making processes on social media, reading and writing with text, images, videos, and emojis combined in a single text (i.e. multimodal texts, Jewitt and Kress 2003 ; Bateman et al. 2017 ; Bezemer and Cowan 2021 ). This multimodal literacy shapes how students learn and use language ( Kress 2003 ; Kress and Rowsell 2018 ). For example, the traditionally separate genre activities of viewing and producing picture sharing, story narration, and commentary making now often occur in integrated genre activities on social platforms (e.g. Instagram, TikTok, Tweets, WhatsApp, and Youtube). Such social practices might increase students’ use of MDM strategies, particularly regarding the combination of narrative and argument discourse modes in students’ writing; online video/picture sharing often accompanies story narration (mainly narrative discourse) and commentary (mainly argument discourse) together (see an example in Figure 1 ).
An example of multimodal text, ‘the face of guilt’. Notes. Shared by DOG@Instagram ( https://www.instagram.com/p/CcD4WjCJIUe/ ) on 8 April 2022. The left of the figure is a photo with the title ‘the face of guilt’, in which a golden retriever is holding a teddy toy with a guilty face. The caption on the photo says, ‘When dad has to go back in the store to pay for the toy he didn’t notice you grabbed and carried out…’ [narrative discourse mode]. The right of the figure shows laughing emojis and textual commentaries from others (e.g. ‘Dogs don’t have rights, so can’t be held accountable for theft’ [argument discourse mode]).
An MDM strategy capitalizes on the complementary strengths of each discourse mode. For example, argumentative writing aims to inform and persuade readers; thus, reasoning is paramount ( Perkins et al. 1991 ; Walton 2005 ). Conversely, narrative recounts a plot of events in a specific way ( tellability , Brunner 1991 ) to be comprehensible, memorable, and persuasive and to arouse emotions ( Labov and Waletzky 2003 ). The detailed particulars of a narrative discourse mode can support the reasoning process and strengthen the claims of the argument ( Bruner 1991 ). Consider the following argument discourse mode with Toulmin’s (2003) argument structures.
Claim: Honest people tend to be more successful. Evidence: They often get more opportunities and make more money than dishonest people. Warrant: Companies try to hire and promote honest people. Backing: Companies do not want their employees to cheat them. Rebuttal: However, some employees successfully hide their dishonesty.
In this example, the writer draws our attention to a claim and its evidence, warrant, backing, and rebuttal ( Toulmin 2003 ). It begins with a claim (honest people tend to be more successful) for which there is possible counter or opposing claims ( debatable claim , Stein and Miller 1993 ). Then, the writer presents observational evidence to support the claim (they often get more opportunities and make more money than dishonest people). This evidence is warranted by another observation (companies try to hire and promote honest people) and backed by a reason (companies do not want their employees to cheat them). Finally, the claim ends with a rebuttal invoking contradictory evidence in favour of an opposing claim (some employees successfully hide their dishonesty).
Next, consider this story of young George Washington and the cherry tree within Labov’s narrative structures ( Labov and Waletzky 2003 ):
Orientation: When George was six years old, he had an axe, and chopped everything in his way. Temporal organization: When George’s dad saw his favourite cherry tree chopped down, he angrily asked, ‘who did it?’ George hesitated but said, ‘I can’t tell a lie, Pa; I cut it.’ To George’s surprise, his dad praised him, ‘Your honesty is worth more than a thousand trees.’ Coda: Being honest is a golden character in young George. He stayed honest and true, eventually becoming the first US president.
In this example, the narrative contains a temporally ordered sequence of situational particulars that answer wh-questions (e.g. who, when, what: George, six years old, chopped down the cherry tree); thus, readers who are familiar with most of the particulars can readily understand it (comprehensible). Readers can store this sequence of particulars in episodic and long-term memories within their cerebral cortex (memorable). The expressed emotional particulars (angrily, hesitated, surprise) arouses the reader’s emotions and motivations, linking their memories stored in the cerebral cortex to their emotions in the amygdala (emotional arousal). Finally, the coda uses elements such as morality (being honest is a golden character in young George) and implications (he stayed honest and true, eventually becoming the first US president) to transmit moral models concerning social situations, thereby promoting social values (persuasion, Deane et al. 2015 ). Thus, the tellability of this narrative coheres with the rhetorical purpose of the argument.
An MDM strategy is a deliberate top-down organization strategy for writing an essay. A discourse strategy that uses both argument and narrative enjoys the rhetorical benefits of the complementary strengths of each. Consider the following example:
[multi-discourse mode strategy with argument followed by narrative] Honest people tend to be more successful. As companies don’t want their employees to cheat them, they try to hire honest people (though some employees successfully hide their dishonesty). Also, honest people are often regarded as trustworthy at work, so they get more opportunities for promotion, compared to dishonest people. It’s like ‘George Washington and the cherry tree’. The six-year-old George stayed honest and true after he chopped down his father’s cherry tree. His father praised his honesty, and he eventually becomes the first US president. [Moral implications:] So, be honest to succeed.
In this MDM strategy example, the argument in the first three sentences is followed by a narrative (the subsequent four sentences), which strengthens the episodic memory effects to make the argument’s reasoning more memorable, comprehensible, emotionally arousing, and persuasive. After the narrative, further argumentation explicitly specifies the narrative’s moral implications (be honest to succeed), making the essay more persuasive ( Labov and Waletzky 2003 ). Therefore, we hypothesized that MDM essays outscore single-discourse essays (H-1):
H-1. Essays with an MDM strategy have higher scores than those with a single-discourse mode strategy.
To date, no published study investigated the relationship between the use of MDM strategy and student essay scores. As discussed above, the MDM essays enjoy the rhetorical benefits of employing the narrative’s situational particulars to make an argument more memorable and more persuasive. Hence, we predict that essays with MDMs have higher scores than those with a single-discourse mode.
Although an MDM strategy can improve the quality of an essay, its disadvantages are the additional time and effort that a student needs to implement it, including selecting and organizing the discourse modes. The writer first selects appropriate discourse modes and content. After choosing to employ, for example, argumentative and narrative discourse modes, the writer must select appropriate content for each mode that fit together. For example, the moral value of the cherry tree story should fit the argumentative claim that ‘honest people tend to be more successful’. Furthermore, the writer considers the macrostructure of the essay, regarding the proportions of the respective discourse modes (primarily argumentative, primarily narrative, or balanced) and where to place the content of each discourse mode: argument followed by narrative, narrative followed by argument, or interleaved. Next, the writer constructs suitable transition(s) between the two discourse modes to maintain the flow of the writing with cohesive devices such as connective words and bridging sentences (e.g. it’s like George Washington and the cherry tree).
To accomplish these additional tasks in MDM writing, writers need better writing skills and typically produce longer essays. Previous studies have shown that longer essays (with more words, H-2a, Kobrin et al. 2007 ; Crossley et al. 2015 ), and essays that used more cohesive devices (such as connective words) often received higher scores (H-2b, McNamara et al. 2010 ). Also, essays with higher syntactic complexity (longer sentences, H-2c) or lexical complexity (lower frequency words, H-2d, and higher lexical diversity, H-2e) often have higher scores ( Crossley and McNamara 2010 ; Crossley et al. 2011 ; McNamara et al. 2013 ). Hence, we hypothesized that, (i) compared to single-discourse mode essays, MDM essays have more words, more cohesive devices, higher syntactic complexity, and greater lexical complexity, and (ii) these attributes mediate the link between MDM essays and higher essay scores (H-2a, b, c, d, e). Additionally, narrative discourses that recount simple, temporally ordered stories may include more concrete words ( narrative particulars; e.g. ‘flowers, perfumes’ as the tokens of ‘gift’, Bruner 1991 : 6). Also, essays by older (often advanced) writers often have more concrete words in their essays ( lexical concreteness , H-2f, Crossley et al. 2011 ).
H-2a. More words mediate the link between the MDM strategy and higher essay scores. H-2b. Proportionally more connective words mediate the link between MDM strategy and higher essay scores. H-2c. Longer mean sentence length mediates the link between MDM strategy and higher essay scores. H-2d. Words that are less common (greater lexical complexity) mediate the link between MDM strategy and higher essay scores. H-2e. Proportionally greater measure of textual lexical diversity (MTLD) mediates the link between MDM strategy and higher essay scores. H-2f. Proportionally more concrete words mediate the link between MDM strategy and higher essay scores.
Participants in this study chose one of two essays with different prompts (picture or text, see Method) which could affect the essay score ( Way et al. 2000 ; Yang et al. 2015 ; Shi et al. 2020 ). Tasks with higher reasoning demands might elicit more lexical and syntactic complexity to yield superior performance (H-3a, cognition hypothesis; Robinson 2001 , 2003 ). As writers must decode spatial, casual, and intentional relations in picture-prompt tasks but not in text-prompt tasks, the higher reasoning demands of the picture prompt may yield more complex writing (e.g. MDM strategies, more words, greater syntactic complexity, greater lexical complexity) and, therefore, better task performance (e.g. higher essay scores). Hence, we proposed hypotheses H-3a and b.
H-3a. Essays using the picture-prompt task receive higher scores than those for the text-prompt task. H-3b. The MDM strategy, the total number of words, and syntactic and lexical complexity mediate the link between prompts and essay scores.
We found no published studies on MDM strategies in writing essays. Our study tested our three sets of hypotheses regarding whether an MDM strategy is related to linguistic complexity and essay scores in a high-stakes language exam.
The data for this study were drawn from previously completed, required examination essays by Grade 12 secondary students. The native Chinese students wrote Chinese language essays during the examination. We randomly sampled 695 anonymous essays with either a text ( n = 332) or picture ( n = 363) argumentative essay prompt. The requirement for both essays was the same: choose one prompt and write a 650-word essay in 1.5 hours ( Figure 2 ).
The text-prompted and picture-prompted writing topics in this study. Notes. The text-prompted task: write an essay entitled ‘Not to be the first and not to be the last either’ to express your opinion of this attitude toward lives. The picture-prompted task: write an essay entitled ‘The sun and the shadow’ according to the pictures.
Grade 12 students in Hong Kong write an essay during a Diploma of Secondary Education exam, and students use these scores on their applications to universities in Hong Kong. Our data consisted of the essays and official essay scores for 695 students. Scores were given by two experienced examination markers who independently assessed each essay on a 7-point rating scale (1, 2, 3, 4, 5, 5*, 5**, ordered from lowest to highest) using a rubric with four dimensions: (i) content, (ii) expression, (iii) structure, and (iv) mechanics and handwriting. If two markers disagreed, a third marker scored the essay and the closest, highest pair of marks was used ( Hong Kong Examinations and Assessment Authority [HKEAA] 2019 ).
Operational definition..
We defined an MDM strategy in an essay as including more than one discourse mode. For the scope of this paper, we only focus on the MDM strategy of employing both narrative and argumentation discourse modes. We coded an argumentative essay as an MDM essay if it has at least one occurrence of a narrative discourse mode that supported the writer’s argument by recounting an event or experience ( Labov and Waletzky 2003 ). We did not code the quantity, proportion, or structures of narrative components.
Two coders coded the essays by dividing them into MDM essays with both narrative and argument discourse modes, versus those with only an argument discourse mode. As anonymous government-employed raters scored the essays in an earlier year, we could not hire them to do additional coding. The first author of this paper coded all 695 essays and divided them into MDM essays with narrative components versus one-discourse mode essays using only argument discourse mode (with no narrative components). The second coder, a senior research assistant with a postgraduate degree in linguistics, coded half of the essays (173 text-prompt essays and 174 picture-prompt essays). For a small effect size of 0.2 and α = 0.05, a sample of 322 is sufficient to yield 95 per cent statistical power, so the second coder coded half of the essays (347) rather than all of them. They worked independently and coded the essays separately. As shown in Table 2 , the two coders showed high inter-rater reliability (99 per cent agreement, Krippendorff’s (2012) α = 0.977). Disagreements were resolved through discussion until they reached a consensus.
Reliability of coding multi-discourse modes (MDM) essays (N = 695)
. | Text-prompt . | Picture-prompt . | Total essays . |
---|---|---|---|
Single coded | 332 | 363 | 695 |
One-discourse mode | 100 | 60 | 160 |
MDM | 232 | 303 | 535 |
Double coded | 173 | 174 | 347 |
One-discourse mode | 52 | 33 | 85 |
MDM | 121 | 138 | 259 |
% Agreement | 100% | 98% | 99% |
Krippendorff’sα | 1.000 | .946 | .977 |
. | Text-prompt . | Picture-prompt . | Total essays . |
---|---|---|---|
Single coded | 332 | 363 | 695 |
One-discourse mode | 100 | 60 | 160 |
MDM | 232 | 303 | 535 |
Double coded | 173 | 174 | 347 |
One-discourse mode | 52 | 33 | 85 |
MDM | 121 | 138 | 259 |
% Agreement | 100% | 98% | 99% |
Krippendorff’sα | 1.000 | .946 | .977 |
Notes. MDM refers to essays that employ the multi-discourse mode strategy.
We used the Chinese Coh-Metrix tool ( Graesser et al. 2011 ; Lu et al. 2017 ) to measure the linguistic complexity of the essays, including the total number of words, syntactic and lexical complexity, and connectedness. Past studies showed that these measures were connected to essay quality ( Crossley and McNamara 2010 ; Crossley et al. 2011 ; McNamara et al. 2013 ). See measures in Table 3 .
Measures used in this study
Variable names . | Measures . |
---|---|
Outcome | |
Essay score | 7-point rating scale (7 = highest) of human-rated scores based on four dimensions: (i) content, (ii) expression, (iii) structure, and (iv) mechanics and handwriting |
Explanatory variables | |
Picture prompt | Picture prompt (versus text prompt; 1 = picture, 0 = text) |
MDM | Coded the multi-discourse mode strategy of narrative and argument modes (1 = MDM, 0 = one-discourse mode) |
Total words | Total number of words in an essay |
% Connectives | Total number of connective words per 1,000 words |
Mean sentence length | Mean number of words per sentence |
Content word frequency | Content word frequency was calculated using the Chinese Coh-Metrix tool. Common words (e.g. bird) have higher frequencies than uncommon words (e.g. pterodactyl). |
MTLD | An index of lexical diversity is not influenced by text length. It is the mean length of sequential word strings in a text that maintains a given type-token ration value ( ) |
Word concreteness | Content word concreteness (versus abstractness) calculated using the Chinese Coh-Metrix |
Variable names . | Measures . |
---|---|
Outcome | |
Essay score | 7-point rating scale (7 = highest) of human-rated scores based on four dimensions: (i) content, (ii) expression, (iii) structure, and (iv) mechanics and handwriting |
Explanatory variables | |
Picture prompt | Picture prompt (versus text prompt; 1 = picture, 0 = text) |
MDM | Coded the multi-discourse mode strategy of narrative and argument modes (1 = MDM, 0 = one-discourse mode) |
Total words | Total number of words in an essay |
% Connectives | Total number of connective words per 1,000 words |
Mean sentence length | Mean number of words per sentence |
Content word frequency | Content word frequency was calculated using the Chinese Coh-Metrix tool. Common words (e.g. bird) have higher frequencies than uncommon words (e.g. pterodactyl). |
MTLD | An index of lexical diversity is not influenced by text length. It is the mean length of sequential word strings in a text that maintains a given type-token ration value ( ) |
Word concreteness | Content word concreteness (versus abstractness) calculated using the Chinese Coh-Metrix |
Notes. All measures were computed using the Traditional Chinese Coh-Metrix except for essay score, picture prompt, and the MDM strategy ( http://cohmetrix.com/ ).
a Mean sentence length was measured by a ratio of total words to sentences ( Cai 2013 ). In Chinese, a word can be one, two, three, or even four characters long.
We ran hierarchical sets of regressions followed by a structural equation model to test whether MDM essays outperformed one-discourse mode essays, controlling for linguistic features. As the prompt type might affect the writing of the essay, we entered it first, before the essay measures. Then, we entered the top-down MDM strategy, followed by holistic features of essay discourse (‘Total number of words’ and ‘% Connectives’). Next, we entered the syntactic variable (‘Mean sentence length’), followed by the smaller lexical complexity variables (‘Content word frequency’, ‘MTLD’, and ‘Word concreteness’). As omitting non-significant variables did not cause omitted variable bias, we safely removed them to increase precision and reduce multicollinearity ( Kennedy 2008 ).
To test for multiple mediation effects simultaneously, we ran a structural equation model via Mplus 7.4 ( Muthén and Muthén 1998–2015 ). To assess the goodness of fit, we used the comparative fit index (CFI), Tucker–Lewis index (TLI), root mean square error approximation (RMSEA), and standardized root mean square residual (SRMR), which minimize Type I and Type II errors under many simulation conditions ( Hu and Bentler 1999 ). We used two-fit thresholds: good (CFI and TLI > 0.95; RMSEA < 0.06; SRMR < 0.08) and moderate (0.90 < CFI and TLI < 0.95;0.06 < RMSEA < 0.10;0.08 < SRMR < 0.10).
Approximately half of the students chose the picture prompt (52 per cent, see Table 4 ). Among these essays, 77 per cent employed the multi-discourse (narrative and argument) modes. The mean essay score was 4.28 ( SD = 2.54, range: 1–7). The mean essay length was 641 words ( SD = 165), close to the maximum of 650 words allowed.
Descriptive statistics and correlation–variance–covariances of variables (N = 695)
Variable . | . | . | 1 . | 2 . | 3 . | 4 . | 5 . | 6 . | 7 . | 8 . | 9 . |
---|---|---|---|---|---|---|---|---|---|---|---|
1.Essay score | 4.28 | 2.54 | 0.26 | 0.40 | 250.76 | −10.76 | −0.83 | −0.29 | 52.91 | 0.09 | |
2.Prompt | 0.52 | 0.50 | 0.20 | 0.03 | −4.30 | −2.15 | −0.09 | −0.03 | 8.29 | 0.06 | |
3.MDM | 0.77 | 0.42 | 0.37 | 0.16 | 23.19 | −1.30 | −0.08 | −0.02 | 5.47 | 0.02 | |
4.Total words | 640.7 | 165.4 | 0.61 | −0.05 | 0.33 | −389.5 | −23.01 | −8.62 | 1,774.7 | 2.48 | |
5.% Connectives | 34.48 | 11.44 | −0.37 | −0.38 | −0.27 | −0.21 | 1.80 | 1.29 | −260.6 | −0.77 | |
6.Mean sentence length | 6.63 | 1.56 | −0.21 | −0.12 | −0.12 | −0.09 | 0.10 | 0.05 | −15.48 | −0.04 | |
7.Content words frequency | 7.29 | 0.25 | −0.46 | −0.21 | −0.17 | −0.21 | 0.45 | 0.14 | −7.16 | 0.00 | |
8.MTLD | 120.9 | 48.0 | 0.43 | 0.35 | 0.27 | 0.22 | −0.47 | −0.21 | −0.60 | 1.84 | |
9.Word concreteness | 3.52 | 0.18 | 0.21 | 0.71 | 0.27 | 0.08 | −0.37 | −0.13 | 0.05 | 0.21 |
Variable . | . | . | 1 . | 2 . | 3 . | 4 . | 5 . | 6 . | 7 . | 8 . | 9 . |
---|---|---|---|---|---|---|---|---|---|---|---|
1.Essay score | 4.28 | 2.54 | 0.26 | 0.40 | 250.76 | −10.76 | −0.83 | −0.29 | 52.91 | 0.09 | |
2.Prompt | 0.52 | 0.50 | 0.20 | 0.03 | −4.30 | −2.15 | −0.09 | −0.03 | 8.29 | 0.06 | |
3.MDM | 0.77 | 0.42 | 0.37 | 0.16 | 23.19 | −1.30 | −0.08 | −0.02 | 5.47 | 0.02 | |
4.Total words | 640.7 | 165.4 | 0.61 | −0.05 | 0.33 | −389.5 | −23.01 | −8.62 | 1,774.7 | 2.48 | |
5.% Connectives | 34.48 | 11.44 | −0.37 | −0.38 | −0.27 | −0.21 | 1.80 | 1.29 | −260.6 | −0.77 | |
6.Mean sentence length | 6.63 | 1.56 | −0.21 | −0.12 | −0.12 | −0.09 | 0.10 | 0.05 | −15.48 | −0.04 | |
7.Content words frequency | 7.29 | 0.25 | −0.46 | −0.21 | −0.17 | −0.21 | 0.45 | 0.14 | −7.16 | 0.00 | |
8.MTLD | 120.9 | 48.0 | 0.43 | 0.35 | 0.27 | 0.22 | −0.47 | −0.21 | −0.60 | 1.84 | |
9.Word concreteness | 3.52 | 0.18 | 0.21 | 0.71 | 0.27 | 0.08 | −0.37 | −0.13 | 0.05 | 0.21 |
Notes . Correlations at 0.075 or higher are significant at p < 0.05; the lower-left triangle contains the correlations, the bold numbers along the diagonal are the variances, and the upper-right triangle contains the covariances.
All results discussed below describe the first entry into the regression, statistically controlling for all previously included variables. Ancillary regressions and statistical tests are available upon request.
Prompt, MDM strategy, total words, sentence length, and lexical complexity variables were related to essay score (see Table 5 ). Compared to text-prompt essays, picture-prompt essays had higher scores on average ( b = 1.048; range: 1–7; see Table 5 , model 1), supporting hypothesis H-3a. MDM essays had higher scores than one-discourse mode essays on average ( b = 2.253; Table 5 , model 2), supporting H-1 (see Table 6 ). Longer essays (more total words) often had higher scores ( b = 0.008; Table 5 , model 3); an essay with 125 more words than another achieved on average a score one level higher (0.008 × 125 = 1), supporting H-2a. Essays with longer sentences than others showed lower scores ( b = −0.196; Table 5 , model 4), contradicting H-2c.
Significant unstandardized parameter coefficients of hierarchical linear regressions predicting essay scores
. | Essay score . | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
. | Model 1 . | Model 2 . | Model 3 . | Model 4 . | Model 5 [final] . | |||||
Explanatory variable . | Prompt . | + MDM . | + Total words . | + Sentence length . | + Lexical complexity . | |||||
Picture prompt | 1.048 (0.189) | *** | 0.755 (0.177) | *** | 1.056 (0.147) | *** | 0.991 (0.147) | *** | 0.386 (0.210) | |
MDM | 2.253 (0.211) | *** | 1.052 (0.187) | *** | 0.993 (0.185) | *** | 0.765 (0.176) | *** | ||
Total words | 0.008 (0.000) | *** | 0.008 (0.000) | *** | 0.007 (0.000) | *** | ||||
Mean sentence length | −0.196 (0.047) | *** | −0.147 (0.044) | ** | ||||||
Content word frequency | −3.158 (0.303) | *** | ||||||||
Word concreteness | 1.165 (0.579) | ** | ||||||||
0.043 | 0.178 | 0.439 | 0.453 | 0.530 |
. | Essay score . | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
. | Model 1 . | Model 2 . | Model 3 . | Model 4 . | Model 5 [final] . | |||||
Explanatory variable . | Prompt . | + MDM . | + Total words . | + Sentence length . | + Lexical complexity . | |||||
Picture prompt | 1.048 (0.189) | *** | 0.755 (0.177) | *** | 1.056 (0.147) | *** | 0.991 (0.147) | *** | 0.386 (0.210) | |
MDM | 2.253 (0.211) | *** | 1.052 (0.187) | *** | 0.993 (0.185) | *** | 0.765 (0.176) | *** | ||
Total words | 0.008 (0.000) | *** | 0.008 (0.000) | *** | 0.007 (0.000) | *** | ||||
Mean sentence length | −0.196 (0.047) | *** | −0.147 (0.044) | ** | ||||||
Content word frequency | −3.158 (0.303) | *** | ||||||||
Word concreteness | 1.165 (0.579) | ** | ||||||||
0.043 | 0.178 | 0.439 | 0.453 | 0.530 |
Notes. Standard errors are in parentheses.
** p < 0.01; *** p < 0.001.
Hypothesis test results
Dimension . | Hypothesis . | Result . |
---|---|---|
The MDM strategy Direct effect | H-1. Essays employing the MDM strategy have higher scores than those employing a single-discourse mode. | Supported |
Indirect effect | H-2. The total number of words and syntactic and lexical complexity mediate the link between the MDM strategy and essay scores. | Supported |
Linguistic performance | ||
Total words | H-2a. Essays with more words achieve higher scores. | Supported |
Connective words | H-2b. Essays with a higher percentage of connective words achieve higher scores. | Not supported |
Sentence length | H-2c. Essays with a longer average sentence length achieve higher scores. | Not supported |
Lexical complexity | H-2d. Essays with more low-frequency content words achieve higher scores. | Supported |
Lexical diversity | H-2e. Essays with a higher MTLD achieve higher scores. | Not supported |
Word concreteness | H-2f. Essays with higher content word concreteness achieve higher scores. | Supported |
Prompt | ||
Direct effect | H-3a. Picture prompt tasks achieve higher scores. | Supported |
Indirect effect | H-3b. The MDM strategy, the total number of words, and syntactic and lexical complexity mediate the effects between prompts and essay scores. | Supported |
Dimension . | Hypothesis . | Result . |
---|---|---|
The MDM strategy Direct effect | H-1. Essays employing the MDM strategy have higher scores than those employing a single-discourse mode. | Supported |
Indirect effect | H-2. The total number of words and syntactic and lexical complexity mediate the link between the MDM strategy and essay scores. | Supported |
Linguistic performance | ||
Total words | H-2a. Essays with more words achieve higher scores. | Supported |
Connective words | H-2b. Essays with a higher percentage of connective words achieve higher scores. | Not supported |
Sentence length | H-2c. Essays with a longer average sentence length achieve higher scores. | Not supported |
Lexical complexity | H-2d. Essays with more low-frequency content words achieve higher scores. | Supported |
Lexical diversity | H-2e. Essays with a higher MTLD achieve higher scores. | Not supported |
Word concreteness | H-2f. Essays with higher content word concreteness achieve higher scores. | Supported |
Prompt | ||
Direct effect | H-3a. Picture prompt tasks achieve higher scores. | Supported |
Indirect effect | H-3b. The MDM strategy, the total number of words, and syntactic and lexical complexity mediate the effects between prompts and essay scores. | Supported |
Lexical complexity variables (content word frequency, word concreteness) were also related to essay score. Essays that used more common words (high-frequency words) had lower scores ( b = −3.158; Table 5 , model 5), supporting H-2d. Essays with more concrete words (instead of abstract words) had higher scores ( b = 1.165; Table 5 , model 3), supporting H-2f. Altogether, these explanatory variables accounted for 53 per cent of the variance in essay scores.
The structural equation model showed a moderate fit (see Figure 3 ; CFI = 0.967, TLI = 0.885, RMSEA = 0.087, SRMR = 0.045, χ 2 [6] = 37.67, p < 0.001, R 2 for Essay score = 0.491). Total words and syntactic- and lexical-complexity variables also mediated the link between prompt and essay score. Students who chose the picture prompt more often wrote essays employing the MDM strategy (+0.15, supporting H-3b), fewer words (−0.11), shorter sentences (−0.10), infrequent words (−0.19), and more concrete words (+0.69). The total number of words and syntactic and lexical complexity also mediated the link between the MDM strategy and essay score, supporting H-2a and H-2d. Students who employed the MDM strategy wrote essays using more words (+0.36), shorter sentences (−0.11), infrequent words (−0.17), and more concrete words (+0.14). All other explanatory variables, interactions, and mediation tests were not significant. Notably, the percentage of connectives and MTLD were not significant.
Standardized parameters of significant explanatory variables in a structural equation model of essay scores. Notes. The corresponding value of each arrow is the standardized parameter coefficient. To simplify the view, non-significant paths are omitted. The solid and dashed arrows indicate positive and negative effects, respectively. A thicker line indicates a larger effect size. * p < 0.05; ** p < 0.01; *** p < 0.001; CFI = 0.967; TLI = 0.885; RMSEA = 0.087; SRMR = 0.045; Χ 2 (6) = 37.67; p < 0.001; R 2 for Essay score = 0.491.
Among these time-limited examination essays, those with longer sentences had lower scores, contradicting H-2c, so we examined several examples of longer versus shorter sentences (see translated Excerpts 1 and 2).
Excerpt 1 . 若求平凡,因第一十分困難,如要做全球首富,可能你工作一生都賺不到現時首富十分之一,那便放棄。(If peace is the desire, because being first is hard, if wanting to become the world’s richest person, maybe you could not earn ten percent of the wealth of the richest person after you work for your whole life, then to give it up). Excerpt 2. 不做第一,最顯而易見的益處固然是避免自己風頭過盛,以致招來禍患。(Not being first has the most obvious benefit of avoiding a high profile, which can otherwise cause a disaster.)
Excerpt 1 is a long sentence from a low-scoring essay. It has five segments separated by commas (a simpler paraphrase is ‘Few people can be first in their fields in society, so living a peaceful life isn’t a bad choice for an ordinary person’). Excerpt 2 is an example of a shorter sentence from a high-scoring essay. It contains fewer words than Excerpt 1, segmented into three sections (theme, benefit, and counter-consequence); it is more concise and easier to understand than the longer sentence. In these examination essays, weaker writers often wrote longer, incoherent sentences, while superior writers wrote concise, coherent ones.
As no published study had examined how an MDM strategy affected essay scores, this was the first study to show that (i) essays written using such a strategy (i.e. combining argumentation and narrative) had higher essay scores and (ii) the total number of words and the lexical and syntactic complexity mediated the link between the MDM strategy and essay score.
Essays with an MDM strategy combining narrative and argument discourse modes had higher essay scores, which aligns with our theoretical model regarding the benefits of integrating multiple discourse modes. Our results are consistent with our claim that integrating narrative discourse with argumentation in an essay enables temporal and situational ‘particulars’ (narrative) to strengthen the argumentative reasoning, making it more memorable, comprehensible, emotionally sustainable, and persuasive ( Labov and Waletzky 2003 ).
Furthermore, the MDM strategy has a direct effect on essay score, even after controlling for the total number of words and lexical and syntactic complexity. This direct effect suggests that the MDM strategy is a distinguishable, separate writing skill. Future studies can test whether writers develop MDM strategies at the later knowledge-crafting stage, in which they show greater awareness of the relations between the author, the text, and its readers ( Kellogg 2008 ; Bereiter and Scardamalia 2013 ).
The total number of words, sentence length, word frequency, and concreteness mediated the positive link between the MDM strategy and essay score. Compared to one-discourse mode essays, the MDM essays often contained more words, which coheres with our hypothesis that MDM essays require more content. Infrequently used words (suggesting more difficult vocabulary) occurred more often in MDM essays than in one-discourse mode essays, which aligns with our hypothesis that greater writing skills (such as mastery of more complex vocabulary) are required to write MDM essays. Furthermore, shorter sentences and concrete words are suggestive of narrative effects ( Bruner 1991 ). Sentences were shorter in the MDM essays (argument and narrative) than in one-discourse mode essays (argument only), which is consistent with the view that narrative essays typically contain simpler sentence structures and, therefore, shorter sentences. Additionally, concrete words occurred more often in the MDM essays than in one-discourse mode essays, which aligns with the view that narrative essays often describe concrete events and, therefore, use many concrete words. We present an example of the MDM strategy in Appendix A ( Supplementary Material ).
Linguistic complexity was also linked to essay scores. Essays with more words or more infrequent words had higher scores than other essays, cohering with the findings of previous studies ( Crossley and McNamara 2010 ; Crossley et al. 2011 ; McNamara et al. 2013 ). In these time-limited examination essays, those with longer sentences had lower scores. Under time pressure, superior writers constructed concise, coherent sentences, while weaker writers constructed longer, incoherent ones. In untimed essays written outside of class, those with greater syntactic complexity achieved higher scores ( McNamara et al. 2010 ). Possibly, writers who have sufficient time to plan/organize their ideas can construct better, longer sentences ( Graham and Perin 2007 ; Graham and Sandmel 2011 ; Graham et al. 2015 ).
Our results showed that the picture prompt was linked to MDM strategy, fewer words, shorter sentence length, more uncommon words, and more concrete words. These results are consistent with the view that different cognitive demands across tasks (e.g. prompts) can activate different prior knowledge to yield greater (or less) complexity in a student’s writing ( Yang et al. 2015 ; Shi et al. 2020 ). They support Robinson’s (2001 , 2003 ) cognition hypothesis: tasks with higher reasoning demands elicit more lexical and syntactic complexity to yield better performance. Hence, task designers and assessors can vary the cognitive demands of a task by using different task prompts in language assessment ( Robinson 2001 , 2003 ). Regardless of the task or prompt, examination essay is a special genre (‘timed-impromptu writing’, Perelman 2014 ) that differs from real-world language use, so essay performance might not generalize to other genre contexts ( Bachman 2002 ).
Although writing is critical to literacy and socioeconomic success in modern society ( National Research Council 2006 ), writing is notoriously difficult to learn and teach ( Bazerman et al. 2018 ). This study contributes a new construct to the relevant literature: an MDM writing strategy. Furthermore, this study shows that among these examination essays, those with multiple discourse modes had higher scores. As this MDM strategy capitalizes on the strengths of each discourse mode to achieve rhetoric purposes, teaching students to effectively use it might improve their writing; future intervention studies can test this idea and collect and analyse qualitative data regarding students’ MDM learning and use processes. For example, researchers can do ethnographies to identify students’ relevant social contexts (online or offline) and understand how they integrate them in their narratives and arguments in an MDM essay; such information can inform educators’ design of lessons about MDM strategies.
This study also has implications for writing assessment. The results show that six variables (prompt, the MDM strategy, total number of words, sentence length, content word frequency, and word concreteness) accounted for 53 per cent of the variance in essay scores, a substantial improvement over past models using only linguistic complexity and discourse cohesion indicators (whose explained variances ranged from 22 to 44 per cent; McNamara et al. 2010 , 2013 ). If future studies show similar, superior results, then later automated essay scoring algorithms can use the MDM strategy to assess essays with greater accuracy. More generally, this study supports Deane’s (2011) call for further investigation of writers’ top-down conceptual strategies and sociocultural knowledge (e.g. schematic, pragmatic, register, and world knowledge). To further advance our understanding of writing and writing strategies, and the relevant instructions and assessment, future studies can examine literacy learning in school contexts and students’ language use in their daily life, especially on social media (social practice perspective).
Our study has limitations concerning the scope of the MDM strategy, languages, student ages, and essay sampling. First, in this paper, we focus only on the combination of narrative and argument discourse modes in argumentative essays. Future studies can study MDM strategies involving more combinations of discourse modes. For example, such studies can study students’ (i) reports that combine report and description discourse modes or (ii) stories that combine narrative and description discourse modes ( Smith 2003 ). As these tasks reflect students’ real-world language uses, they are particularly worth studying for pedagogical purposes. Second, we only analysed essays written in Chinese by native Chinese speakers; future studies can examine essays in other languages and those written by natives and/or second-language writers. Third, we only studied essays written by upper secondary school students; future studies can investigate essays written by students of different ages, which might illuminate age constraints concerning the use of MDM strategies. Researchers can also conduct qualitative studies on whether advanced writers tend to use more MDM strategies, and what interventions for teaching MDM strategies can succeed. Fourth, our essay sample only included two essay topics from a high-stakes examination; future studies can examine more topics and concerning different stakes (e.g. freewriting in a journal, class assignment, creative writing competition). Fifth, our data do not include students’ school lessons (such as whether teachers taught MDM strategies), textbooks, or home learning environment, so future studies can collect such data to test a richer explanatory model—for example, using controlled experiments to investigate whether short-term or long-term teaching of MDM strategies improves students’ essay writing or literacy development.
The authors would like to thank the editors and the anonymous reviewers for their insightful comments and suggestions. They also want to thank Tsz Wing Choi, Yik Ting Choi, Sze Ming Lam, Alice Liu, Ka Yee Mak, Magdalena Mo Ching Mok, and Alex Morakhovski for helping with the data collection, coding, and analysis.
Bachman , L. F. 2002 . ‘Some reflections on task-based language performance assessment,’ Language Testing 19 / 4 : 453 – 76 .
Google Scholar
Barton , D. 2007 . Literacy: An Introduction to the Ecology of Written Language (2nd edn). Blackwell .
Google Preview
Barton , D. and C. Lee . 2013 . Language Online: Investigating Digital Texts and Practices . Routledge .
Barton , D. and D. Potts . 2013 . ‘Language learning online as a social practice,’ TESOL Quarterly 47 / 4 : 815 – 20 .
Bateman , J. , J. Wildfeuer , and T. Hiippala . 2017 . Multimodality: Foundations, Research and Analysis–A Problem-Oriented Introduction . Walter de Gruyter GmbH & Co KG .
Bazerman , C. , et al. .. 2018 . The Lifespan Development of Writing . National Council of Teachers of English .
Bereiter , C. and M. Scardamalia . 2013 . The Psychology of Written Communication . Routledge .
Bezemer , J. and K. Cowan . 2021 . ‘Exploring reading in social semiotics: Theory and methods,’ Education 3-13 49 / 1 : 107 – 18 .
Biber , D. , B. Gray , and S. Staples . 2016 . ‘Predicting patterns of grammatical complexity across language exam task types and proficiency levels,’ Applied Linguistics 37 : 639 – 68 .
Bruner , J. 1991 . ‘The narrative construction of reality,’ Critical Inquiry 18 : 1 – 21 .
Cai , X. Q. 2013 . ‘ The development and application of textual and lexical frequency indices and analysis system for children’, Unpublished master’s thesis, National Taichung University of Education, available at . https://hdl.handle.net/11296/hm53fz
Crossley , S. and D. S. McNamara . 2010 . ‘Cohesion, coherence, and expert evaluations of writing proficiency’ in R. Catrambone and S. Ohlsson (eds): Proceedings of the 32nd Annual Conference of the Cognitive Science Society. Cognitive Science Society , pp. 984 – 89 .
Crossley , S. A. , et al. . 2011 . ‘The development of writing proficiency as a function of grade level: A linguistic analysis,’ Written Communication 28 : 282 – 311 .
Crossley , S. A. , K. Kyle , and D. S. McNamara . 2015 . ‘To aggregate or not? Linguistic features in automatic essay scoring and feedback systems,’ Journal of Writing Assessment 8 / 1 : 1 – 19 .
Deane , P. 2013 . ‘On the relation between automated essay scoring and modern views of the writing construct,’ Assessing Writing 18 : 7 – 24 .
Deane , P. 2011 . Writing Assessment and Cognition (ETS Research Report RR–11–14). Educational Testing Service . Retrieved from https://www.ets.org/research/policy_research_reports/publications/report/2011/ingy.html
Deane , P. , et al. . 2015 . Key Practices in the English Language Arts (ELA): Linking Learning Theory, Assessment, and Instruction , Research Report No. RR–15–17. Educational Testing Service .
DOG@Instagram (April 8, 2022 ). The Face of Guilt. Instagram , available at https://www.instagram.com/p/CcD4WjCJIUe/
Ehrenhalt , A. 1998 . ‘ Hijacking the rulebook’ . New York Times , December 20, Section 4: 13 .
Ferretti , R. P. and S. Graham . 2019 . ‘Argumentative writing: Theory, assessment, and instruction,’ Reading and Writing 32 : 1345 – 57 .
Graesser , A. C. , D. S. McNamara , and J. M. Kulikowich . 2011 . ‘Coh-Metrix: Providing multilevel analyses of text characteristics,’ Educational Researcher 40 : 223 – 34 .
Graham , S. 2018 . ‘A revised writer (s)-within-community model of writing,’ Educational Psychologist 53 : 258 – 79 .
Graham , S. , K. R. Harris , and T. Santangelo . 2015 . ‘Research-based writing practices and the common core: Meta-analysis and meta-synthesis,’ The Elementary School Journal 115 : 498 – 522 .
Graham , S. and D. Perin . 2007 . ‘A meta-analysis of writing instruction for adolescent students,’ Journal of Educational Psychology 99 : 445 – 76 .
Graham , S. and K. Sandmel . 2011 . ‘The process writing approach: A meta-analysis, ’ The Journal of Educational Research 104 : 396 – 407 .
Hayes , J. R. and L. S. Flower . 1986 . ‘Writing research and the writer,’ American Psychologist 41 : 1106 – 13 .
Hong Kong Examinations and Assessment Authority (HKEAA) . 2019 . Introduction to HKDSE Core Subjects. HKSAR, available at https://www.hkeaa.edu.hk/en/hkdse/introduction/ Accessed 23 Sep. 2020 .
Hu , L. T. and P. M. Bentler . 1999 . ‘Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives,’ Structural Equation Modeling: A Multidisciplinary Journal 6 : 1 – 55 .
Jewitt , C. and G. Kress . 2003 . Multimodal Literacy . Peter Lang .
Kellogg , R. T. 2008 . ‘Training writing skills: A cognitive developmental perspective,’ Journal of Writing Research 1 : 1 – 26 .
Kennedy , P. 2008 . Guide to Econometrics . Wiley-Blackwell .
Kobrin , J. L. , H. Deng , and E. J. Shaw . 2007 . ‘Does quantity equal quality: The relationship between length of response and scores on the SAT essay,’ Journal of Applied Testing Technology 8 / 1 : 1 – 15 .
Kress , G. 2003 . Literacy in the New Media Age . Routledge .
Kress , G. and J. Rowsell . 2018 . ‘Literacy as a social practice: new realities and new models’ In D. Bloome , M. L. Castanheira , C. Leung and J. Rowsell (eds): Re-Theorizing Literacy Practices: Complex Social and Cultural Context . Routledge , pp. 30 – 49 .
Krippendorff , K. 2012 . Content Analysis . Sage .
Labov , W. and J. Waletzky . 2003 . Narrative Analysis: Oral Versions of Personal Experience . University of Washington Press , pp. 74 – 104 .
Lu , Y. J. , B. C. Kuo , and K. C. Pai . 2017 . ‘Developing Chinese automated essay scoring model to assess college students’ essay quality’ in X. Hu , T. Barnes , A. Hershkovitz and L. Paquette (eds): Proceedings of the 10th International Conference on Educational Data Mining , pp. 430 – 432 .
McCarthy , P. M. and S. Jarvis . 2010 . ‘MTLD, vocd-D, and HD-D: A validation study of sophisticated approaches to lexical diversity assessment,’ Behavior Research Methods 42 : 381 – 92 .
McNamara , D. S. , et al. . 2015 . ‘A hierarchical classification approach to automated essay scoring,’ Assessing Writing 23 : 35 – 59 .
McNamara , D. S. , S. A. Crossley , and P. M. McCarthy . 2010 . ‘Linguistic features of writing quality,’ Written Communication 27 : 57 – 86 .
McNamara , D. S. , S. A. Crossley , and R. Roscoe . 2013 . ‘Natural language processing in an intelligent writing strategy tutoring system,’ Behavior Research Methods 45 : 499 – 515 .
Muthén , L. K. and B. O. Muthén . 1998–2015 . Mplus User’s Guide , 7th edn. CA , Muthén and Muthén .
National Center for Education Statistics . 2012 . The Nation’s Report Card: Writing 2011 (NCES 2012–470) . Institute of Education Sciences, U.S. Department of Education .
National Research Council . 2006 . Measuring Literacy: Performance Levels for Adults . National Academies Press .
Perelman , L. 2014 . ‘When “the state of the art” is counting words,’ Assessing Writing 21 : 104 – 11 .
Perkins , D. N. , M. Farady , and B. Bushey . 1991 . ‘Everyday reasoning and the roots of intelligence’ in J. F. Voss , D. N. Perkins , and J. W. Segal (eds): Informal Reasoning and Education . Lawrence Erlbaum Associates , pp. 83 – 105 .
Robinson , P. 2003 . ‘The cognition hypothesis, task design, and adult task-based language learning,’ Second Language Studies 21 : 45 – 105 .
Robinson , P. 2001 . ‘Task complexity, task difficulty, and task production: Exploring interactions in a componential framework,’ Applied Linguistics 22 : 27 – 57 .
Shi , B. , L. Huang , and X. Lu . 2020 . ‘Effect of prompt type on test-takers’ writing performance and writing strategy use in the continuation task,’ Language Testing 37 : 361 – 88 .
Smith , C. S. 2001 . ‘Discourse modes: aspectual entities and tense interpretation,’ Cahiers de Grammaire 26 : 183 – 206 .
Smith , C. S. 2003 . Modes of Discourse: The Local Structure of Texts , Vol. 103 . Cambridge University Pres s.
Smith . C. S. 2013 . The Parameter of Aspect , Vol. 43 . Springer Science and Business Media .
Sommers , N. 1980 . ‘Revision strategies of student writers and experienced writers,’ College Composition and Communication 31 : 378 – 87 .
Stein , N. and C. A. Miller . 1993 . ‘A theory of argumentative understanding: Relationships among position preference, judgments of goodness, memory and reasoning,’ Argumentation 7 : 183 – 204 .
Toulmin , S. E. 2003 . The Uses of Argument . Cambridge University Press .
Walton , D. 2005 . Fundamentals of Critical Argumentation . Cambridge University Press .
Way , D. P. , E. G. Joiner , and M. A. Seaman . 2000 . ‘Writing in the secondary foreign language classroom: The effects of prompts and tasks on novice learners of French,’ The Modern Language Journal 84 : 171 – 84 .
Yang , W. , X. Lu , and S. C. Weigle . 2015 . ‘Different topics, different discourse: Relationships among writing topic, measures of syntactic complexity, and judgments of writing quality,’ Journal of Second Language Writing 28 : 53 – 67 .
Month: | Total Views: |
---|---|
January 2023 | 261 |
February 2023 | 179 |
March 2023 | 204 |
April 2023 | 135 |
May 2023 | 136 |
June 2023 | 82 |
July 2023 | 51 |
August 2023 | 66 |
September 2023 | 97 |
October 2023 | 112 |
November 2023 | 103 |
December 2023 | 85 |
January 2024 | 66 |
February 2024 | 60 |
March 2024 | 426 |
April 2024 | 360 |
May 2024 | 358 |
June 2024 | 286 |
July 2024 | 158 |
August 2024 | 98 |
Citing articles via, looking for your next opportunity.
Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide
Sign In or Create an Account
This PDF is available to Subscribers Only
For full access to this pdf, sign in to an existing account, or purchase an annual subscription.
The rise of an academic theory and its obsession with Israel
O n October 7 , Hamas killed four times as many Israelis in a single day as had been killed in the previous 15 years of conflict. In the months since, protesters have rallied against Israel’s retaliatory invasion of Gaza, which has killed tens of thousands of Palestinians. But a new tone of excitement and enthusiasm could be heard among pro-Palestinian activists from the moment that news of the attacks arrived, well before the Israeli response began. Celebrations of Hamas’s exploits are familiar sights in Gaza and the West Bank, Cairo and Damascus; this time, they spread to elite college campuses, where Gaza-solidarity encampments became ubiquitous this past spring. Why?
The answer is that, long before October 7, the Palestinian struggle against Israel had become widely understood by academic and progressive activists as the vanguard of a global battle against settler colonialism, a struggle also waged in the United States, Canada, Australia, and other countries created by European settlement. In these circles, Palestine was transformed into a standard reference point for every kind of social wrong, even those that seem to have no connection to the Middle East.
One of the most striking things about the ideology of settler colonialism is the central role played by Israel, which is often paired with the U.S. as the most important example of settler colonialism’s evils. Many Palestinian writers and activists have adopted this terminology. In his 2020 book, The Hundred Years’ War on Palestine , the historian Rashid Khalidi writes that the goal of Zionism was to create a “white European settler colony.” For the Palestinian intellectual Joseph Massad, Israel is a product of “European Jewish Settler-Colonialism,” and the “liberation” referred to in the name of the Palestine Liberation Organization is “liberation from Settler-Colonialism.”
Western activists and academics have leaned heavily on the idea. Opposition to building an oil pipeline under a Sioux reservation was like the Palestinian cause in that it “makes visible the continuum of systems of subjugation and expropriation across liberal democracies and settler-colonial regimes.” When the city of Toronto evicted a homeless encampment from a park, it was like Palestine because both are examples of “ethnic cleansing” and “colonial ‘domicide,’ making Indigenous people homeless on their homelands.” Health problems among Native Americans can be understood in terms of Palestine, because the “hyper-visible Palestine case … provides a unique temporal lens for understanding settler colonial health determinants more broadly.” Pollution, too, can be understood through a Palestinian lens, according to the British organization Friends of the Earth, because Palestine demonstrates that “the world is an unequal place” where “marginalised and vulnerable people bear the brunt of injustice.”
Although Israel fails in obvious ways to fit the model of settler colonialism, it has become the standard reference point because it offers theorists and activists something that the United States does not: a plausible target. It is hard to imagine America or Canada being truly decolonized, with the descendants of the original settlers returning to the countries from which they came and Native peoples reclaiming the land. But armed struggle against Israel has been ongoing since it was founded, and Hamas and its allies still hope to abolish the Jewish state “between the river and the sea.” In the contemporary world, only in Israel can the fight against settler colonialism move from theory to practice.
T he concept of settler colonialism was developed in the 1990s by theorists in Australia, Canada, and the U.S., as a way of linking social evils in these countries today—such as climate change, patriarchy, and economic inequality—to their origin in colonial settlement. In the past decade, settler colonialism has become one of the most important concepts in the academic humanities, the subject of hundreds of books and thousands of papers, as well as college courses on topics such as U.S. history, public health, and gender studies.
Read: The curious rise of settler colonialism and Turtle Island
For the academic field of settler-colonial studies, the settlement process is characterized by European settlers discovering a land that they consider “terra nullius,” the legal property of no one; their insatiable hunger for expansion that fills an entire continent; and the destruction of Indigenous peoples and cultures. This model, drawn from the history of Anglophone colonies such as the U.S. and Australia, is regularly applied to the history of Israel even though it does not include any of these hallmarks.
When modern Zionist settlement in what is now Israel began in the 1880s, Palestine was a province of the Ottoman empire, and after World War I, it was ruled by the British under a mandate from the League of Nations. Far from being “no one’s land,” Jews could settle there only with the permission of an imperial government, and when that permission was withdrawn—as it fatefully was in 1939, when the British sharply limited Jewish immigration on the eve of the Holocaust—they had no recourse. Far from expanding to fill a continent, as in North America and Australia, the state of Israel today is about the size of New Jersey. The language, culture, and religion of the Arab peoples remain overwhelmingly dominant: 76 years after Israel was founded, it is still the only Jewish country in the region, among 22 Arab countries, from Morocco to Iraq.
Most important, the Jewish state did not erase or replace the people already living in Palestine, though it did displace many of them. Here the comparison between European settlement in North America and Jewish settlement in Israel is especially inapt. In the decades after Europeans arrived in Massachusetts, the Native American population of New England declined from about 140,000 to 10,000, by one estimate . In the decades after 1948, the Arab population of historic Palestine more than quintupled, from about 1.4 million to about 7.4 million. The persistence of the conflict in Israel-Palestine is due precisely to the coexistence of two peoples in the same land—as opposed to the classic sites of settler colonialism, where European settlers decimated Native peoples.
In the 21st century, the clearest examples of ongoing settler colonialism can probably be found in China. In 2023, the United Nations Human Rights office reported that the Chinese government had compelled nearly 1 million Tibetan children to attend residential schools “aimed at assimilating Tibetan people culturally, religiously and linguistically.” Forcing the next generation of Tibetans to speak Mandarin is part of a long-term effort to Sinicize the region, which also includes encouraging Han Chinese to settle there and prohibiting public displays of traditional Buddhist faith.
China has mounted a similar campaign against the Uyghur people in the northwestern province of Xinjiang. Since 2017, more than 1 million Uyghur Muslims have been detained in what the Chinese government calls vocational training centers, which other countries describe as detention or reeducation camps. The government is also seeking to bring down Uyghur birth rates through mass sterilization and involuntary birth control.
These campaigns include every element of settler colonialism as defined by academic theorists. They aim to replace an existing people and culture with a new one imported from the imperial metropole, using techniques frequently described as genocidal in the context of North American history. Tibet’s residential schools are a tool of forced assimilation, like the ones established for Native American children in Canada and the United States in the 19th century. And some scholars of settler colonialism have drawn these parallels, acknowledging, in the words of the anthropologist Carole McGranahan, “that an imperial formation is as likely to be Chinese, communist, and of the twentieth or twenty-first centuries as it is to be English, capitalist, and of the eighteenth or nineteenth centuries.”
Yet Tibet and Xinjiang—like India’s rule in Kashmir, and the Indonesian occupation of East Timor from 1975 to 1999—occupy a tiny fraction of the space devoted to Israel-Palestine on the mental map of settler-colonial studies. Some of the reasons for this are practical. The academic discipline mainly flourishes in English-speaking countries, and its practitioners usually seem to be monolingual, making it necessary to focus on countries where sources are either written in English or easily available in translation. This rules out any place where a language barrier is heightened by strict government censorship, like China. Just as important, settler-colonial theorists tend to come from the fields of anthropology and sociology rather than history, area studies, and international relations, where they would be exposed to a wider range of examples of past and present conflict.
But the focus on Israel-Palestine isn’t only a product of the discipline’s limitations. It is doctrinal. Academics and activists find adding the Israeli-Palestinian conflict to other causes powerfully energizing, a way to give a local address to a struggle that can otherwise feel all too abstract. The price of collapsing together such different causes, however, is that it inhibits understanding of each individual cause. Any conflict that fails to fit the settler-colonial model must be made to fit.
I srael also fails to fit the model of settler colonialism in another key way: It defies the usual division between foreign colonizers and Indigenous people. In the discourse of settler colonialism, Indigenous peoples aren’t simply those who happen to occupy a territory before Europeans discovered it. Rather, indigeneity is a moral and spiritual status, associated with qualities such as authenticity, selflessness, and wisdom. These values stand as a reproof to settler ways of being, which are insatiably destructive. And the moral contrast between settler and indigene comes to overlap with other binaries—white and nonwhite, exploiter and exploited, victor and victim.
Until recently, Palestinian leaders preferred to avoid the language of indigeneity, seeing the implicit comparison between themselves and Native Americans as defeatist. In an interview near the end of his life, in 2004, PLO Chair Yasser Arafat declared, “We are not Red Indians.” But today’s activists are more eager to embrace the Indigenous label and the moral valences that go with it, and some theorists have begun to recast Palestinian identity in ecological, spiritual, and aesthetic terms long associated with Native American identity. The American academic Steven Salaita has written that “Palestinian claims to life” are based in having “a culture indivisible from their surroundings, a language of freedom concordant to the beauty of the land.” Jamal Nabulsi of the University of Queensland writes that “Palestinian Indigenous sovereignty is in and of the land. It is grounded in an embodied connection to Palestine and articulated in Palestinian ways of being, knowing, and resisting on and for this land.”
This kind of language points to an aspect of the concept of indigeneity that is often tacitly overlooked in the Native American context: its irrationalism. The idea that different peoples have incommensurable ways of being and knowing, rooted in their relationship to a particular landscape, comes out of German Romantic nationalism. Originating in the early 19th century in the work of philosophers such as Johann Gottlieb Fichte and Johann Gottfried Herder, it eventually degenerated into the blood-and-soil nationalism of Nazi ideologues such as Richard Walther Darré, who in 1930 hymned what might be called an embodied connection to Germany: “The German soul, with all its warmness, is rooted in its native landscape and has, in a sense, always grown out of it … Whoever takes the natural landscape away from the German soul, kills it.”
For Darré, this rootedness in the land meant that Germans could never thrive in cities, among the “rootless ways of thinking of the urbanite.” The rootless urbanite par excellence, for Nazi ideology, was of course the Jew. For Salaita, the exaltation of Palestinian indigeneity leads to the very same conclusion about “Zionists,” who usurp the land but can never be vitally rooted in it: “In their ruthless schema, land is neither pleasure nor sustenance. It is a commodity … Having been anointed Jewish, the land ceases to be dynamic. It is an ideological fabrication with fixed characteristics.”
In this way, anti-Zionism converges with older patterns of anti-Semitic and anti-Jewish thinking. It is true, of course, that criticism of Israel is not inherently anti-Semitic. Virtually anything that an Israeli government does is likely to be harshly criticized by many Israeli Jews themselves. But it is also true that anti-Semitism is not simply a matter of personal prejudice against Jews, existing on an entirely different plane from politics. The term anti-Semitism was coined in Germany in the late 19th century because the old term, Jew hatred , sounded too instinctive and brutal to describe what was, in fact, a political ideology—an account of the way the world works and how it should be changed.
Wilhelm Marr, the German writer who popularized the word, complained in his 1879 book, The Victory of Judaism Over Germanism , that “the Jewish spirit and Jewish consciousness have overpowered the world.” That spirit, for Marr, was materialism and selfishness, “profiteering and usury.” Anti-Semitic political parties in Europe attacked “Semitism” in the same way that socialists attacked capitalism. The saying “Anti-Semitism is the socialism of fools,” used by the German left at this time, recognized the structural similarity between these rival worldviews.
The identification of Jews with soulless materialism made sense to 19th-century Europeans because it translated one of the oldest doctrines of Christianity into the language of modern politics. The apostle Paul, a Jew who became a follower of Jesus, explained the difference between his old faith and his new one by identifying Judaism with material things (the circumcision of the flesh, the letter of the law) and Christianity with spiritual things—the circumcision of the heart, a new law “written not with ink but with the Spirit of the living God, not on tablets of stone but on tablets of human hearts.”
Simon Sebag Montefiore: The decolonization narrative is dangerous and false
Today this characterization of Jews as stubborn, heartless, and materialistic is seldom publicly expressed in the language of Christianity, as in the Middle Ages, or in the language of race, as in the late 19th century. But it is quite respectable to say exactly the same thing in the language of settler colonialism. As the historian David Nirenberg has written, “We live in an age in which millions of people are exposed daily to some variant of the argument that the challenges of the world they live in are best explained in terms of ‘Israel,’” except that today, Israel refers not to the Jewish people but to the Jewish state.
When those embracing the ideology of settler colonialism think about political evil, Israel is the example that comes instinctively to hand, just as Jews were for anti-Semitism and Judaism was for Christianity. Perhaps the most troubling reactions to the October 7 attacks were those of college students convinced that the liberation of Palestine is the key to banishing injustice from the world. In November 2023, for instance, Northwestern University’s student newspaper published a letter signed by 65 student organizations—including the Rainbow Alliance, Ballet Folklórico Northwestern, and All Paws In, which sends volunteers to animal shelters—defending the use of the slogan “From the river to the sea, Palestine will be free.” This phrase looks forward to the disappearance of any form of Jewish state between the Mediterranean and the Jordan, but the student groups denied that this entails “murder and genocide.” Rather, they wrote, “When we say from the river to the sea, Palestine will be free, we imagine a world free of Islamophobia, antisemitism, anti-Blackness, militarism, occupation and apartheid.”
As a political program, this is nonsensical. How could dismantling Israel bring about the end of militarism in China, Russia, or Iran? How could it lead to the end of anti-Black racism in America, or anti-Muslim prejudice in India? But for the ideology of settler colonialism, actual political conflicts become symbolic battles between light and darkness, and anyone found on the wrong side is a fair target. Young Americans today who celebrate the massacre of Israelis and harass their Jewish peers on college campuses are not ashamed of themselves for the same reason that earlier generations were not ashamed to persecute and kill Jews—because they have been taught that it is an expression of virtue.
This essay is adapted from Adam Kirsch’s new book, On Settler Colonialism: Ideology, Violence, and Justice .
When you buy a book using a link on this page, we receive a commission. Thank you for supporting The Atlantic.
More Stories
When Poetry Could Define a Life
The Smartest Man Who Ever Lived
IMAGES
COMMENTS
Narrative discourse aims to entertain, engage, and transport readers or listeners into a fictional world through vivid descriptions, compelling characters, and an engaging plot. Examples of narrative discourse in literature include novels, short stories, epics, and folktales.
Interactive example of a narrative essay. An example of a short narrative essay, responding to the prompt "Write about an experience where you learned something about yourself," is shown below. Hover over different parts of the text to see how the structure works. Narrative essay example.
gations, it is the centerpiece of the study of narrative, for in attempting to define the forms and figures of narrative discourse Genette must deal with all the complex relations between the narrative and the story it tells. The structures and codes which Barthes and Todorov studied must be taken up and organized
'Narrative Theory' is an online introduction to classical structuralist narratological analysis. The third section deals with the narrative articulation of time, taking as a guideline Gérard Genette's theory in 'Narrative Discourse', modified as required. Outline: 1. The Generation of Story Time. 2. Fabula time. 3. Story time: order. 4.
For example, academic essays employ argumentative discourse to persuade readers about the truth of an overarching thesis. ... Examples of Discourse in Literature. 1. William Shakespeare, Romeo and Juliet. Shakespeare's classic tragedy of star-crossed teenage lovers opens with a narrative discourse: Two households, both alike in dignity,In ...
Interpretive approach: Discourse analysis is an interpretive approach, meaning that it seeks to understand the meaning and significance of language use from the perspective of the participants in a particular discourse. Emphasis on reflexivity: Discourse analysis emphasizes the importance of reflexivity, or self-awareness, in the research process.
Use clear and concise language throughout the essay. Much like the descriptive essay, narrative essays are effective when the language is carefully, particularly, and artfully chosen. Use specific language to evoke specific emotions and senses in the reader. The use of the first person pronoun 'I' is welcomed. Do not abuse this guideline!
Narrative Discourse: An Essay in Method. Narrative Discourse. : Gérard Genette. Cornell University Press, 1980 - Language Arts & Disciplines - 285 pages. Genette uses Proust's Remembrance of Things Past as a work to identify and name the basic constituents and techniques of narrative. Genette illustrates the examples by referring to other ...
Narrative is understood by structuralists. to be a form of communication. The real. authors communicate a story (the formal. content element of narrative) by. discourse (the formal expression ...
Narrative Essay Definition. Writing a narrative essay is a unique form of storytelling that revolves around personal experiences, aiming to immerse the reader in the author's world. It's a piece of writing that delves into the depths of thoughts and feelings. In a narrative essay, life experiences take center stage, serving as the main substance of the story. It's a powerful tool for writers ...
Views: 36,041. Narrative discourse is a type of written or verbal communication that involves narration, or in other words, telling a story. This is one of the classic types of discourse that helps people to identify different modes of communication and different kinds of functions for speaking or writing. The narrative discourse is common in ...
In fact, it's organized into three categories: Written discourse: Composed of written works like essays, blog posts, and books. Spoken discourse: Shared through speech, like presentations, vlogs, and oral reports. Civil discourse: Spoken or written words characterized by its inclusion of multiple participants, all of whom engage on a level ...
4 pages / 1843 words. A discourse community is a conglomerate of individuals that share common knowledge and use the same rhetoric to communicate ideas within a specialized topic. The financial sector is a prime example of an extremely exclusive discourse community that is very difficult to become a part...
Critical discourse analysis (or discourse analysis) is a research method for studying written or spoken language in relation to its social context. It aims to understand how language is used in real life situations. When you conduct discourse analysis, you might focus on: The purposes and effects of different types of language.
Originally, it has roots in the Latin language. The term assumes slightly different meanings in different contexts.In literature, discourse means speech or writing, normally longer than sentences, which deals with a certain subject formally. In other words, discourse is the presentation of language in its entirety, while performing an intellectual inquiry in a particular area or field, such as ...
Order Narrative Discourse Essay Sample or use it for FREE. 1(888)606-0029 1(888)676-5101 Live Chat. Order. menu. Home ... action performers. In a narrative discourse, events are chronologically arranged and use a first or third person. Examples of narrative discourse include folk stories, historical events, myths and personal experience stories
Genette illustrates the examples by referring to other literary works. His systemic theory of narrative deals with the structure of fiction, including fictional devices that go unnoticed and whose implications fulfill the Western narrative tradition Translation of: Discours du récit, a portion of the 3d vol. of the author's Figures, essais
Discourse community essay is an essential part of academic writing that requires students to explore a specific group's communication methods and practices. To write a successful discourse community essay, you need to understand the group's language, values, and beliefs. Here are some tips on how to write a discourse community essay that ...
For example, the narrative discourse mode is paramount in the narrative fiction genre and the argumentative-commentary discourse mode is common in the editorial genre; however, a text can ... Fourth, our essay sample only included two essay topics from a high-stakes examination; future studies can examine more topics and concerning different ...
Essays marked with a * received a distinction. * Analyzing and raising students' awareness of textual patterns in authentic texts: Mohammad Umar Farooq. Written Text Analysis: Gregory S. Hadley. *Show an analysis of the whole text in terms of the main underlying text pattern. Identify the signals that indicate this pattern David Evans.
Narrative discourse, or the ability to recall and tell an orderly and continuous account of an event or a series of events, is a vital aspect of social... Essays Topics
This essay is adapted from Adam Kirsch's new book, On Settler Colonialism: Ideology, Violence, and Justice. On Settler Colonialism: Ideology, Violence, and Justice By Adam Kirsch