Multilingual fiction is often discussed through reader inclusion and exclusion, a binary that can obscure the performative work of glosses, naming, typography, voicing, and narrated translation. This article asks how multilingual signals organize relations among speakers, audiences, and forms of authority when prose is approached as verbal art capable of vocal realization. It compares public-domain short prose by Sui Sin Far, Onoto Watanna, Zitkala-Ša, and Rabindranath Tagore. Performance-oriented close reading follows code-switching, glossing, naming, address, syntax, typography, and metalinguistic explanation, recording each signal's relation to imagined speaking, listening, and interpretation. The analysis argues that untranslated language can create relation rather than simply exclude readers, that glosses disclose imagined hierarchies between audiences, and that narrated translation can stage negotiations between familial and institutional authority. By distinguishing linguistic presence from narrative and performative function, the article develops a relational account of multilingual thresholds for literary performance and audience address. Its conclusions remain specific to the selected texts and translations; claims about actual performers, listeners, readership, linguistic competence, or diasporic reception require evidence beyond textual analysis.

The interpretation developed here examines how short fiction uses multilingual signals to allocate belonging, effort, voice, and interpretive responsibility. Its corpus comprises public-domain prose by Sui Sin Far, Onoto Watanna, Zitkala-Ša, and Rabindranath Tagore, read at the level of code-switching, glossing, naming, address, narrated translation, and cues that condition vocal delivery. The central claim is that untranslated language can create connection rather than simply exclude readers. That proposition matters to performance-oriented criticism because multilingual form makes the labor of speaking, hearing, and understanding perceptible within written evidence. Attention therefore falls on passages where more than one language is signaled through vocabulary, syntax, typography, naming, or metalinguistic explanation.

The question is narrower than a general history of multilingual performance. Within these texts, a multilingual threshold is identified when a linguistic signal changes the relation among speaker, addressee, implied audience, or subsequent interpretation. The article treats prose as verbal art with performable features, not as evidence that a particular staged realization occurred. This distinction keeps the performance claims proportionate to the textual evidence.

The evidence is synthesized in three findings. First, untranslated language can create connection rather than simply excluding readers. Second, glosses can reveal an imagined hierarchy between audiences by distributing explanation unevenly. Third, narrated translation can become a scene in which familial and institutional authority are negotiated. Together these findings show that multilingual address is neither decorative texture nor an identity label. It is a formal arrangement of voices, access, timing, and response.

The relation between textual detail and performative possibility supplies a second commitment. Bauman's account of verbal art as performance directs attention to the organization of speaking, roles, audience, and event rather than to language as an isolated inventory of forms [1]. Taylor's distinction between archive and repertoire clarifies how written traces can inform embodied transmission without becoming equivalent to performance [2].

A narrow comparative question gives Languages at the Doorway its reach. Instead of asking what the selected works mean as wholes, the analysis asks how multilingual passages regulate access, vocal position, and interpretive responsibility under different formal conditions. The repeated question makes the cases comparable while retaining their historically and materially different answers. A resistant instance can revise one inference without forcing every text into the same performance model.

The article defines multilingual thresholds through consequences that can be traced in the text. A formal event belongs to the analysis only when it alters expectation, access, authority, memory, vocal alignment, or the range of plausible readings. This criterion connects the three claims while preserving disagreement about individual instances. It also supplies a stopping rule: when the asserted change cannot be located in the linguistic, typographic, narrative, or metalinguistic organization of a passage, the observation is described without assigning it a stronger performative or historical significance.

The research question is therefore how untranslated language, glosses, and narrated translation organize relations among speakers, implied audiences, and authorities, and how those relations condition the performability of the prose. The objective is to compare formal consequences without inferring performer intention or audience reception. The article next establishes its critical context and method, develops three findings, discusses their relation, and closes with limitations and a conclusion.

Multilingual literary study prevents code-switching from being treated as a transparent sign of identity. Yildiz examines the pressure exerted by the monolingual paradigm on modern literary cultures [3]. Sommer shows how bilingual aesthetics can make partial comprehension, linguistic mismatch, and interpretive effort active conditions of reading [4]. These perspectives complement attention to who is positioned to speak, hear, understand, or remain outside an exchange.

Paratextual theory clarifies how glosses and explanations frame linguistic difference and regulate access to a text [5]. Ricoeur's account of narrative temporality supports attention to the way later events can reconfigure the significance of earlier ones [6]. Said's critique of representation makes questions of cultural authority inseparable from the historical conditions under which voices are framed and interpreted [7]. Fish's theory of interpretive communities cautions against treating meaning as the product of an isolated reader alone [8]. Moretti's work on scale, although methodologically distinct from close reading, helps mark the limit between conclusions drawn from a small corpus and claims about larger literary populations [9]. Here these approaches join Bauman's performance-centered verbal art and Taylor's embodied repertoire, allowing textual address to be studied as an invitation to voicing and listening without claiming a specific performance event [1,2].

The primary materials are Sui Sin Far's Mrs. Spring Fragrance [10], Onoto Watanna's short story “Miss Lily and Miss Chrysanthemum” [11], Zitkala-Ša's American Indian Stories [12], and Rabindranath Tagore's The Hungry Stones, and Other Stories [13]. Cases are used for conceptual contrast rather than statistical representation. Each presents a distinct configuration of multilingual address and permits comparison through code-switching, glossing, naming, direct address, typography, narrated translation, and cues that can affect pacing or vocal differentiation in oral interpretation.

The principal protocol is performance-oriented comparative discourse analysis. The codebook records the linguistic signal, whether it is translated or glossed, the speaking or narrating position, the implied addressee, the surrounding typographic or syntactic cue, and the narrative consequence. A shift of voice, pace, emphasis, or listener alignment is treated as a textual affordance rather than proof of historical delivery.

The evidentiary scale of Languages at the Doorway is deliberately local. The procedure records the position, recurrence, interruption, and outcome of multilingual signals before assigning interpretive weight. A repeated feature may retain its appearance while changing function, whereas a gloss, restatement, or change of speaker can redirect the sequence. The article therefore withholds claims about intention, reception, prevalence, or measured audience response because the present method supplies no empirical evidence for them.

Because multilingual thresholds depend on placement, source variation matters. Omissions or changes in pagination, editing, typography, or translation can alter the documented relation among a linguistic cue, its surrounding voice, and its later interpretation. Version-specific observations are kept separate from conclusions that hold across the cited work. This practice narrows some claims, but it also prevents a polished textual format from concealing uncertainty about the precise formal context in which a multilingual signal operates.

The first pattern is that untranslated language can create relation rather than simply excluding readers. The claim concerns organization, not inventory. A multilingual expression may withhold immediate semantic access while still marking affiliation, intimacy, distance, rhythm, or a change in the relation between speaking and listening positions. The performance-oriented question is what non-translation asks a reader or performer to do with voice, pause, emphasis, and uncertainty.

The resulting form is recursive. Later textual evidence may revise what initially appeared settled, so an untranslated expression can acquire a different function when a later gloss, translation, response, or narrative consequence returns to it. This recursion explains why multilingual address can remain coherent without becoming predictable and why placement matters as much as lexical content.

The relevant unit is therefore a connected sequence rather than a detached example. A passage is followed from preparation through occurrence to effect. This allows the analysis to distinguish structural emphasis from sheer frequency and to show why a rare linguistic event may carry more interpretive force than a frequently repeated one. Without a demonstrable shift in access, expectation, memory, voice, or relation, the stronger claim about a multilingual threshold must be narrowed.

Sui Sin Far's Mrs. Spring Fragrance is approached through this sequence-based test. The analysis restricts itself to passages in which linguistic difference is signaled through wording, naming, syntax, typography, or explicit explanation and asks how the surrounding narrative positions speakers and addressees. The performance-oriented question is whether the cue reorganizes the imagined act of address: for example, whether it changes who appears to possess linguistic authority, who must infer meaning, or how a line would be differentiated in oral realization.

The proposition is evaluated through placement rather than keyword recurrence. An observation functions as evidence only when its documented position changes the interpretation that follows. A multilingual cue is not treated as confirmation merely because it resembles the thesis; its sequence must demonstrate a change in relation, access, or vocal alignment. This constraint keeps the analysis from converting cultural difference into a performance effect by assertion alone.

A close reading must therefore identify the relevant linguistic occurrence, the expectation active at that moment, and the outcome that distinguishes it from background recurrence. The case supports the argument only when those points can be located in the cited text. General resemblance to another case is insufficient; the episode must show why a multilingual threshold is needed to explain the organization of address.

Onoto Watanna's “Miss Lily and Miss Chrysanthemum” provides the second comparison. Its premise places linguistic non-understanding between two sisters, making multilingual difference part of an interpersonal scene rather than a detachable ornament [11]. The governing question is whether explanation, naming, or subsequent mediation redistributes access between the characters and the audiences positioned around them. This is treated as a structured problem of speaking and listening.

The Watanna case changes the scale of comparison from isolated vocabulary to the distribution of access and response. Recurrence matters only when a later instance transfers interpretive force to a different voice, register, participant, or position. That distributive test supplies the pressure behind the second claim: a gloss can expose an imagined hierarchy between audiences when explanation is offered to one position while another is left to infer, wait, or remain partially outside the exchange.

The principal counter-reading treats such distribution as ordinary narrative convention with no special bearing on multilingual performance. The article answers only at the level of the text: it asks whether a transfer of access or emphasis produces a subsequent effect that convention alone leaves unexplained. If it does not, the proposition contracts. Hierarchy is therefore treated as an event of textual organization only when the documented redistribution of linguistic access changes what follows.

Zitkala-Ša's American Indian Stories gives priority to revision over initial recognition. The interpretive test is whether narrated translation or linguistic mediation becomes a scene in which familial and institutional authority are negotiated. Its mixture of autobiographical narrative, stories, and commentary requires care because shifts in voice or linguistic authority cannot be generalized across the volume [12].

The value of this comparison is retrospective. A later passage can be reread against an earlier linguistic encounter, allowing the analysis to distinguish repetition from revision. If the later event merely restates the initial one, the proposition receives little support. If it changes the status of what the reader or implied listener has already encountered, the case makes visible how a multilingual threshold can acquire compositional memory and how a potential oral realization would need to register that changed status.

Verification therefore requires attention to both the initial treatment of linguistic difference and the later moment that alters its status. The interval between them matters because a retrospective reading depends on what could have been remembered, anticipated, or misrecognized. Recording this sequence prevents the later interpretation from being projected backward as though it had been available from the start.

The final case, Tagore's The Hungry Stones, and Other Stories, functions as a resistant comparison [13]. Because the volume contains English translations produced by several hands, its multilingual evidence is mediated at the level of translation as well as narration. That condition tests how far the claim can travel before translation history, genre, or editorial mediation requires qualification.

The resistant role of the Tagore volume also limits the article's vocabulary. If a translation changes the availability of the linguistic feature under discussion, the comparison names that difference before proposing an analogy. The case may therefore support only part of the claim that untranslated or partially translated language can create connection. Such partial support is informative because it identifies the point at which the performance-oriented interpretation ceases to explain the evidence and another historical or textual account must take over.

The second claim holds that glosses can reveal an imagined hierarchy between audiences. This pattern shifts analysis from isolated form toward distribution. A gloss is not treated simply as helpful information; its position can disclose which addressee is presumed to need explanation, which voice is authorized to provide it, and whether the explanation interrupts or sustains the rhythm of an exchange. In oral interpretation, those same formal differences may condition emphasis, pacing, and the relation between quoted speech and explanatory narration.

A third pattern suggests that narrated translation can become a scene in which familial and institutional authority are negotiated. Return, correction, reframing, and delayed effect are distinct analytical events. The relevant question is whether translation merely supplies semantic equivalence or whether the act of explaining, renaming, repeating, or interpreting relocates authority among narrator, character, institution, and implied audience.

The findings collectively form a sequence of their own. The first identifies a process that directs attention. The second shows that access and authority are distributed. The third explains how later events may revise that distribution. This sequence supports the central thesis that multilingual address is historically variable, formally organized, and potentially performative because it structures relations of speaking and listening before any actual performance is documented.

Revision is neither correction for its own sake nor automatic proof of complexity. To sustain the third claim, a later phrase, explanation, response, or narrative action must reorganize a prior inference about linguistic difference. The comparison therefore specifies what changes and what remains stable. This balance allows multilingual thresholds to retain continuity without becoming predictable and marks the point at which ambiguity is structured by the text rather than supplied by commentary.

The discussion applies the three propositions only to the selected material. A resistant case is useful because it separates parallel from explanation. Evidence that complicates the proposition that untranslated language can create interaction rather than simply exclude readers requires a narrower claim, not dismissal. The same test applies to the proposed hierarchy of glosses and to scenes of narrated translation. Coherence among cases counts only after their differences in genre, translation, voice, and textual mediation have been described.

The cases further indicate that multilingual form can make the labor of speaking, listening, and understanding visible and ethically consequential. This result does not require agreement about every instance of multilingualism. It requires only that the interpretation identify the formal mechanism by which access or authority is redistributed and that the claim remain at the level supported by the text.

The contribution is methodological as well as interpretive. Performance studies provides a useful vocabulary for treating verbal art as organized action, but the distinction between textual affordance and documented performance must remain explicit. A written cue may invite vocal differentiation, pacing, gesture, or audience alignment; it cannot establish performer embodiment or audience reception.

Comparison also requires asymmetry. The four cases need not display multilingual thresholds with equal clarity, and the least cooperative case carries special weight. It tests whether the three claims name connected operations or merely repeat the article's vocabulary. By allowing a claim to contract when the evidence resists it, the analysis turns variation into a condition of interpretation rather than a problem to be edited away.

The comparative gain comes from holding the question steady while allowing the formal answer to vary. Each case is asked how a multilingual threshold is prepared, allocated, and revised through linguistic, typographic, narrative, or metalinguistic cues. Because those cues may operate differently by genre, translation history, or narrative position, surface parallelism is not sufficient evidence. Formally dissimilar instances can perform comparable relational work, while apparently similar instances can require different interpretations.

The final stage describes uncertainty rather than concealing it. When two hypotheses account for a passage with comparable force, the analysis does not select one merely to produce closure. Where a reading remains unresolved, the missing edition, translation history, passage-level locator, or contextual source is identified as the kind of evidence needed to distinguish the alternatives. This practice gives the argument a clear evidentiary boundary and prevents confidence of tone from standing in for support.

The primary boundary of the claim is that translation and edition differences can affect multilingual features. The source-based account does not establish reception, intention, prevalence, performer behavior, or causal effects. Its performing-arts relevance therefore lies in the formal organization of verbal address and in the textual affordances available to oral interpretation, recitation, or staged reading, not in undocumented claims about historical performance practice.

For teaching and public interpretation, the article suggests a three-part practice: identify the linguistic feature, describe the relation that gives it force, and state the historical or performative claim at the level the evidence permits. The sequence keeps multilingual thresholds visible without converting complexity into a slogan. It also makes missing editions, translation records, or reception documents explicit research limits rather than silent assumptions.

The guiding question concerned how short fiction uses multilingual signals to allocate belonging, effort, voice, and interpretive responsibility. Through performance-oriented comparative discourse analysis, the article argued that untranslated language can create relation rather than simply excluding readers; that glosses can reveal an imagined hierarchy between audiences; and that narrated translation can become a scene in which familial and institutional authority are negotiated. These claims converge on a relational account of multilingual address in which written form can organize potential acts of speaking and listening without being conflated with documented performance.

The comparison indicates that multilingual form makes the labor of understanding visible and can condition how voice, pacing, explanation, and audience alignment are imagined in oral or staged interpretation. Its principal performing-arts connection is the treatment of multilingual prose as verbal art whose formal cues have performance consequences without licensing claims about actual performer choices or audience responses.

Conflict of Interest. The authors declare that they have no financial or non-financial conflicts of interest relevant to this work.

Funding. The authors declare that this research received no external funding.

Data and Materials Availability. No new data were created or analysed in this study, and therefore data sharing is not applicable.

Ethics and Consent. This study did not involve human participants, animals, or identifiable personal data; therefore, ethics approval and informed consent were not required.

Author Contributions. All authors contributed to the conception, development, analysis, writing, and revision of the manuscript and approved the final version and author order.

Generative Artificial Intelligence. No substantive generative artificial-intelligence assistance was used in the scientific content, analysis, results, or conclusions of this work.

Version and Provenance. This manuscript represents the authors' final approved version, and any subsequent Version of Record should transparently document its publication and production history.

1. [1] Richard Bauman. “Verbal Art as Performance.” American Anthropologist, 77(2), 290–311, 1975. doi:10.1525/aa.1975.77.2.02a00030.
2. [2] Diana Taylor. The Archive and the Repertoire: Performing Cultural Memory in the Americas. Duke University Press, 2003.
3. [3] Yasemin Yildiz. Beyond the Mother Tongue: The Postmonolingual Condition. Fordham University Press, 2012.
4. [4] Doris Sommer. Bilingual Aesthetics: A New Sentimental Education. Duke University Press, 2004.
5. [5] Gérard Genette. Paratexts: Thresholds of Interpretation. Translated by Jane E. Lewin, Cambridge University Press, 1997.
6. [6] Paul Ricoeur. Time and Narrative, Volume 1. Translated by Kathleen McLaughlin and David Pellauer, University of Chicago Press, 1984.
7. [7] Edward W. Said. Orientalism. Pantheon Books, 1978.
8. [8] Stanley Fish. Is There a Text in This Class? The Authority of Interpretive Communities. Harvard University Press, 1980.
9. [9] Franco Moretti. Distant Reading. Verso, 2013.
10. [10] Sui Sin Far. Mrs. Spring Fragrance. A. C. McClurg & Co., 1912.
11. [11] Onoto Watanna. “Miss Lily and Miss Chrysanthemum: The Love Story of Two Japanese Girls in Chicago.” Ladies' Home Journal, August 1903, pp. 11–12.
12. [12] Zitkala-Ša. American Indian Stories. Hayworth Publishing House, 1921.
13. [13] Rabindranath Tagore. The Hungry Stones, and Other Stories. Translated from the Bengali by various writers, Macmillan, 1916.