Semantic Drift Research Articles

ТОЛКОВАТЕЛИ «ТРУДНЫХ СЛОВ» И СЕМАНТИЧЕСКИЙ ДРЕЙФ В АНГЛИИ XVII–XVIII ВВ.

Изменения в семантическом наполнении средств языковой выразительности служат характерным маркером общественных трансформаций. Причиной могут быть социальные, культурно-религиозные, мировоззренческие и иные структурные изменения, происходящие в ареале бытования языка. Революционные события английской общественной жизни середины XVII в. позволяют предположить, что изменения произошли и в языковой семантике. «Усложнение» английского языка за счет переводов текстов латинских авторов, публикации путешественниками слов чужеземцев, распространение неологизмов и «лексических инноваций» за счет усложнения структуры научного познания – все это требовало кодификации и синтагматизации языка. В статье рассматриваются особенности первых английский словарей «трудных слов». На примере сопоставления словарей Г. Кокерэма и словаря С. Джонсона продемонстрированы изменения, произошедшие в семантическом наполнении слов в Англии XVII в. Делается вывод, что в середине XVIII в. стремление составителей толковых словарей зафиксировать окончательное значение слов наталкивается на изменчивость языка и культурных норм, вынуждая признать утопичность проектов по созданию унифицированных и не требующих дальнейших исправлений словарей. Changes in the semantics of words serve as a characteristic marker of mental transformations. The reason may be social, cultural, religious, ideological and other structural changes occurring in the area where the language exists. The revolutionary events of English social life in the mid-17th century suggest that characteristic transformations also occurred in the semantics of the language. The “complication” of the English language through translations of texts by Latin authors, the publication by travelers of the words of strangers, the spread of neologisms and “lexical innovations” due to the complication of the structure of scientific knowledge - all this required the codification and syntagmatization of the language. The article discusses the features of the first English dictionaries of “hard words”. Using the example of a comparison of the dictionaries of H. Cockeram and the later dictionary of Samuel Johnson, the changes that occurred in the semantic content of words in England in the 17th century are demonstrated. It is concluded that in the middle of the 18th century, the desire of compilers of explanatory dictionaries to fix the final meaning of words ran into the variability of language and cultural norms. They have to admit the utopian nature of projects to create unified dictionaries that do not require further corrections.

Концептуальная инженерия: идея и проблемное поле

The article analyzes the ideas and methods of conceptual engineering, its goals, problem field and opportunities. Conceptual engineering as a philosophical di­rection arises at the intersection of descriptive and normative approaches, repre­senting a revisionist strategy. Emphasizing the correlation between science and social life, the developers of the idea draw attention to the need for a conscious attitude to conceptual representation and terminological choice. Particular atten­tion is paid to the analysis of the name of the direction: the engineering perspec­tive is associated with the idea of ​​managing the scientific process. Theoretical and methodological origins of conceptual engineering are considered: the ideas of J. Dewey on the development of “conscious thinking”, R. Carnap on explica­tion, S. Haslanger on the importance of a critical attitude towards significant so­cial theoretical constructs. It is shown that conceptual engineering as a revi­sionist epistemological project is aimed at changing specific forms of social interaction through the revision of current ideas about the world and about one­self. Given the paramount importance of language in organizing everyday intel­lectual practices, conceptual engineering focuses on concepts and their meaning. In essence, it represents another attempt to eliminate linguistic ambiguity, but no longer with the help of a special symbolic language, but by manipulating mean­ing and sense. The problems of semantic modifications of concepts associated with semantic drift, linguistic intuition and communication, knowledge transfer are analyzed

Consistent representation joint adaptive adjustment for incremental zero-shot learning

Zero-shot learning aims to recognize objects from novel concepts through the model trained on seen class data and assisted by the semantic descriptions. Though it breaks the serious reliance on training data, it still fails to deal with the sequential streaming data in the open world. In this paper, we focus on the incremental zero-shot learning (IZSL) where the seen data arrives in the form of task sequence. In each incremental task, we only have access to the seen data of current task. The IZSL methods aim to depict the characteristics of current seen classes and avoid forgetting the previous ones, in the meantime, learn to generalize to the unseen classes. We summarize the challenges in IZSL as semantic drift which is further divided into task-recency bias and seen-class bias. To solve these issues, we propose a novel IZSL method termed as CRAA. Specifically, CRAA constructs consistent representations with satisfactory discrimination and generalization for all the seen and unseen classes. Based on these representations, CRAA learns a prototype classifier with a novel adaptive adjustment strategy to alleviate the task-recency bias and seen-class bias. Note that CRAA only needs a limited memory footprint to store the fixed scale model, and meets the demands of both memory restriction and data security in industry. We have conducted extensive experiments to evaluate our method on three widely used datasets. The results prove our method is superior to all the compared methods with significant improvements. Code is available at: https://github.com/changniu54/CRAA-Master.

Keystrokes: A practical exploration of semantic drift in timed word association tasks.

This study investigates the phenomena of semantic drift through the lenses of language and situated simulation (LASS) and the word frequency effect (WFE) within a timed word association task. Our primary objectives were to determine whether semantic drift can be identified over the short time (25 seconds) of a free word association task (a predicted corollary of LASS), and whether more frequent terms are generated earlier in the process (as expected due to the WFE). Respondents were provided with five cue words (tree, dog, quality, plastic and love), and asked to write as many associations as they could. We hypothesized that terms generated later in the task (fourth time quartile, the last 19-25 seconds) would be semantically more distant (cosine similarity) from the cue word than those generated earlier (first quartile, the first 1-7 seconds), indicating semantic drift. Additionally, we explored the WFE by hypothesizing that earlier generated words would be more frequent and less diverse. Utilizing a dataset matched with GloVe 300B word embeddings, BERT and WordNet synsets, we analysed semantic distances among 1569 unique term pairs for all cue words across time. Our results supported the presence of semantic drift, with significant evidence of within-participant, semantic drift from the first to fourth time (LASS) and frequency (WFE) quartiles. In terms of the WFE, we observed a notable decrease in the diversity of terms generated earlier in the task, while more unique terms (greater diversity and relative uniqueness) were generated in the 4th time quartile, aligning with our hypothesis that more frequently used words dominate early stages of a word association task. We also found that the size of effects varied substantially across cues, suggesting that some cues might invoke stronger and more idiosyncratic situated simulations. Theoretically, our study contributes to the understanding of LASS and the WFE. It suggests that semantic drift might serve as a scalable indicator of the invocation of language versus simulation systems in LASS and might also be used to explore cognition within word association tasks more generally. The findings also add a temporal and relational dimension to the WFE. Practically, our research highlights the utility of word association tasks in understanding semantic drift and the diffusion of word usage over a sub-minute task, arguably the shortest practically feasible timeframe, offering a scalable method to explore group and individual changes in semantic relationships, whether via the targeted diffusion of influence in a marketing campaign, or seeking to understand differences in cognition more generally. Possible practical uses and opportunities for future research are discussed.

KGRED: Knowledge-graph-based rule discovery for weakly supervised data labeling

In weakly supervised learning, labeling rules can automatically label data to train models. However, due to insufficient prior knowledge, rule discovery often suffers from semantic drift. Since misclassified rules are generated from wrongly matched sentences, the sentences matched by rules shift from the target labels to other labels. It is worth noting that rules do not exist in isolation. The multi-dimensional semantic associations among rules can impose semantic constraints for rule generation, as well as enrich the semantic information of rules for rule matching. Therefore, we propose a Knowledge-Graph-based RulE Discovery method (KGRED), which can leverage the multi-dimensional semantic associations among rules to alleviate semantic drift in rule discovery. Specifically, to decrease misclassified rules, we design a label-aware rule generation approach to attentively propagate prior knowledge from seed rules to candidate rules based on rule KG. To reduce wrongly-matched sentences, we present a cross-attention-based semantic matching mechanism to refine the semantic information of sentences while enriching that of rules. Moreover, we propose an inconsistency-directed active learning strategy to verify rules that perform inconsistently in rule generation and matching. Experiments on two public datasets prove that KGRED can achieve at least 5.1 % gain in F1 score compared to state-of-the-art methods.

Hybrid model for extractive single document summarization: utilizing BERTopic and BERT model

Extractive text summarization has been a popular research area for many years. The goal of this task is to generate a compact and coherent summary of a given document, preserving the most important information. However, current extractive summarization methods still face several challenges such as semantic drift, repetition, redundancy, and lack of coherence. A novel approach is presented in this paper to improve the performance of an extractive summarization model based on bidirectional encoder representations from transformers (BERT) by incorporating topic modeling using the BERTopic model. Our method first utilizes BERTopic to identify the dominant topics in a document and then employs a BERT-based deep neural network to extract the most salient sentences related to those topics. Our experiments on the cable news network (CNN)/daily mail dataset demonstrate that our proposed method outperforms state-of-the-art BERT-based extractive summarization models in terms of recall-oriented understudy for gisting evaluation (ROUGE) scores, which resulted in an increase of 32.53% of ROUGE-1, 47.55% of ROUGE-2, and 16.63% of ROUGE-L when compared to baseline BERT-based extractive summarization models. This paper contributes to the field of extractive text summarization, highlights the potential of topic modeling in improving summarization results, and provides a new direction for future research.

Transfer learning method for specific emitter identification based on pseudo‐labelling and meta‐learning

AbstractSpecific emitter identification (SEI) represents a prominent research direction within the electronic countermeasures domain aimed at discerning carrier identity attributes by analysing subtle radar characteristics. At present, most established SEI techniques assume that both the source and target domain (TD) data adhere to the same distribution. However, this assumption is invalidated by semantic drift which frequently occurs between TD and source domain (SD) samples owing to variations in the collection environment, equipment, and other factors. Considering the aforementioned challenges, this article introduces a transfer learning approach for SEI to leverage pseudo‐label integration within the framework of meta‐learning. This approach employs the bispectral perimeter integral for extracting emitter signal features to construct a feature extractor and basic learner based on CNN13. To label and filter the TD samples, the proposed approach utilises the multiple pseudo‐label serial filtering mechanism, which comprises positive and negative pseudo‐labelling strategies, label uncertainty prediction methods, and hard sample filtering strategies. Ultimately, to address algorithmic real‐time requirements, the labelled TD samples are integrated into the feature extractor and learner of the SD through meta‐learning to facilitate the transfer of TD features to the SD training model. Experimental validation conducted on a real radar dataset demonstrated that the proposed algorithm significantly enhances identification accuracy, exhibiting an improvement from approximately 50% to approximately 90%. Furthermore, the algorithm exhibits a short runtime and robust adaptability, effectively catering to the demands of practical scenarios.

Sparsify dynamically expandable network via variational dropout

This paper develops a new method for lifelong learning referred to as Sparsify Dynamically Expandable Network (SDEN) via Variational Dropout, which explores a sparse model while preserving the performance. Dynamically Expandable Network (DEN) can learn a sequence of tasks via performing network retraining, network expansion by adding only the necessary neurons, and network split to effectively prevent semantic drift in an online manner. To overcome point estimation of parameters in DEN, Bayesian Compression for DEN is developed under the Bayesian framework. However, this method demands more time for model training and testing. To improve the model efficiency, we propose SDEN under the efficient sparse learning framework. We validate our SDEN in the lifelong learning scenarios with multiple frequently used benchmarks, on which it can obtain comparable classification accuracy, and less training and testing time compared with the comparison methods. Furthermore, our method can also learn a more sparse network structure which means fewer network parameters.

A survey of text summarization: Techniques, evaluation and challenges

This paper explores the complex field of text summarization in Natural Language Processing (NLP), with particular attention to the development and importance of semantic understanding. Text summarization is a crucial component of natural language processing (NLP), which helps to translate large amounts of textual data into clear and understandable representations. As the story progresses, it demonstrates the dynamic transition from simple syntactic structures to sophisticated models with semantic comprehension. In order to effectively summarize, syntactic, semantic, and pragmatic concerns become crucial, highlighting the necessity of capturing not only grammar but also the context and underlying meaning. It examines the wide range of summarization models, from conventional extractive techniques to state-of-the-art tools like pre-trained models. Applications are found in many different fields, demonstrating how versatile summarizing techniques are. Semantic drift and domain-specific knowledge remain obstacles, despite progress. In the future, the study predicts developments like artificial intelligence integration and transfer learning, which motivates academics to investigate these prospects for advancement. The approach, which is based on the PRISMA framework, emphasizes a methodical and open literature review. The work attempts to further natural language processing (NLP) and text summarization by combining various research findings and suggesting future research directions in this dynamic subject.

Pois: proprietà discorsive e interazionali di un marcatore discorsivo portoghese tradotto in italiano

"Pois: discursive and interactional properties of a Portuguese discourse marker translated into Italian. In this paper we intend to present the study of the Portuguese discursive marker pois to emphasise the various discursive and interactional possibilities. This discursive connector is included by Lopes e Carrilho (2020) among the explanatory discursive markers: the lexical element takes the form of an operator with anaphoric value that proposes the part of discourse that follows it as an explanation of the part of discourse that precedes it. Part of this semantic value is also found in the pois with an interactional function of response. This phenomenon could be explained by considering the process of semantic drift that sees the expansion of the discursive space of use of a lexical element. To present the various areas of use of the discursive marker pois, written and oral language data will be examined. In addition, a contrastive analysis with Italian will be proposed based on the possible translations that this discourse marker can have in this language in order to broaden the contrastive studies between the two Romance languages. Keywords: discourse markers; Portuguese, Italian; pois; linguistic interaction. "

CL3: Generalization of Contrastive Loss for Lifelong Learning.

Lifelong learning portrays learning gradually in nonstationary environments and emulates the process of human learning, which is efficient, robust, and able to learn new concepts incrementally from sequential experience. To equip neural networks with such a capability, one needs to overcome the problem of catastrophic forgetting, the phenomenon of forgetting past knowledge while learning new concepts. In this work, we propose a novel knowledge distillation algorithm that makes use of contrastive learning to help a neural network to preserve its past knowledge while learning from a series of tasks. Our proposed generalized form of contrastive distillation strategy tackles catastrophic forgetting of old knowledge, and minimizes semantic drift by maintaining a similar embedding space, as well as ensures compactness in feature distribution to accommodate novel tasks in a current model. Our comprehensive study shows that our method achieves improved performances in the challenging class-incremental, task-incremental, and domain-incremental learning for supervised scenarios.

Graph Clustering with High-Order Contrastive Learning.

Graph clustering is a fundamental and challenging task in unsupervised learning. It has achieved great progress due to contrastive learning. However, we find that there are two problems that need to be addressed: (1) The augmentations in most graph contrastive clustering methods are manual, which can result in semantic drift. (2) Contrastive learning is usually implemented on the feature level, ignoring the structure level, which can lead to sub-optimal performance. In this work, we propose a method termed Graph Clustering with High-Order Contrastive Learning (GCHCL) to solve these problems. First, we construct two views by Laplacian smoothing raw features with different normalizations and design a structure alignment loss to force these two views to be mapped into the same space. Second, we build a contrastive similarity matrix with two structure-based similarity matrices and force it to align with an identity matrix. In this way, our designed contrastive learning encompasses a larger neighborhood, enabling our model to learn clustering-friendly embeddings without the need for an extra clustering module. In addition, our model can be trained on a large dataset. Extensive experiments on five datasets validate the effectiveness of our model. For example, compared to the second-best baselines on four small and medium datasets, our model achieved an average improvement of 3% in accuracy. For the largest dataset, our model achieved an accuracy score of 81.92%, whereas the compared baselines encountered out-of-memory issues.

Inherit With Distillation and Evolve With Contrast: Exploring Class Incremental Semantic Segmentation Without Exemplar Memory.

As a front-burner problem in incremental learning, class incremental semantic segmentation (CISS) is plagued by catastrophic forgetting and semantic drift. Although recent methods have utilized knowledge distillation to transfer knowledge from the old model, they are still unable to avoid pixel confusion, which results in severe misclassification after incremental steps due to the lack of annotations for past and future classes. Meanwhile data-replay-based approaches suffer from storage burdens and privacy concerns. In this paper, we propose to address CISS without exemplar memory and resolve catastrophic forgetting as well as semantic drift synchronously. We present Inherit with Distillation and Evolve with Contrast (IDEC), which consists of a Dense Knowledge Distillation on all Aspects (DADA) manner and an Asymmetric Region- wise Contrastive Learning (ARCL) module. Driven by the devised dynamic class-specific pseudo-labelling strategy, DADA distils intermediate-layer features and output-logits collaboratively with more emphasis on semantic-invariant knowledge inheritance. ARCL implements region- wise contrastive learning in the latent space to resolve semantic drift among known classes, current classes, and unknown classes. We demonstrate the effectiveness of our method on multiple CISS tasks by state-of-the-art performance, including Pascal VOC 2012, ADE20K and ISPRS datasets. Our method also shows superior anti-forgetting ability, particularly in multi-step CISS tasks.

Cluster-Guided Contrastive Graph Clustering Network

Benefiting from the intrinsic supervision information exploitation capability, contrastive learning has achieved promising performance in the field of deep graph clustering recently. However, we observe that two drawbacks of the positive and negative sample construction mechanisms limit the performance of existing algorithms from further improvement. 1) The quality of positive samples heavily depends on the carefully designed data augmentations, while inappropriate data augmentations would easily lead to the semantic drift and indiscriminative positive samples. 2) The constructed negative samples are not reliable for ignoring important clustering information. To solve these problems, we propose a Cluster-guided Contrastive deep Graph Clustering network (CCGC) by mining the intrinsic supervision information in the high-confidence clustering results. Specifically, instead of conducting complex node or edge perturbation, we construct two views of the graph by designing special Siamese encoders whose weights are not shared between the sibling sub-networks. Then, guided by the high-confidence clustering information, we carefully select and construct the positive samples from the same high-confidence cluster in two views. Moreover, to construct semantic meaningful negative sample pairs, we regard the centers of different high-confidence clusters as negative samples, thus improving the discriminative capability and reliability of the constructed sample pairs. Lastly, we design an objective function to pull close the samples from the same cluster while pushing away those from other clusters by maximizing and minimizing the cross-view cosine similarity between positive and negative samples. Extensive experimental results on six datasets demonstrate the effectiveness of CCGC compared with the existing state-of-the-art algorithms. The code of CCGC is available at https://github.com/xihongyang1999/CCGC on Github.

Detecting interdisciplinary semantic drift for knowledge organization based on normal cloud model

To reduce the conceptual ambiguity in interdisciplinary knowledge organization systems (KOSs) and enhance interdisciplinary KOS management, this paper proposes a framework for interdisciplinary semantic drift (ISD) detection based on the normal cloud model (NCM). In this framework, we first analyze the features of interdisciplinary concepts and propose a novel interdisciplinary concept extraction method based on cross-discipline statistical information. Secondly, the high-performance knowledge representation model NCM is adopted to represent each interdisciplinary concept with uncertainty, and then a new ISD degree calculation method is proposed based on the similarity cloud algorithm. Thirdly, to identify the direction of ISD after the degree calculation, we propose an ISD direction identification method according to the theory of knowledge potential energy (KPE). Fourthly, based on the above procedure, we propose an ISD detection algorithm to identify and visualize the ISD process. Finally, we evaluate the proposed framework on the concept of “information entropy” and compare the performance with three baselines. Experimental results demonstrate that our framework outperforms[ all the baselines, and the result is comparable to experts’ judgments (0.808 on Spearman correlation, p<0.001). The research indicates the meaning of an interdisciplinary concept will drift from the high KPE discipline to the low KPE discipline as long as interdisciplinary knowledge potential differences (KPD) exist between these two related disciplines. We further identify three key factors that affect the degree of ISD: the length of the discipline chain of an interdisciplinary concept transfer, the number of source disciplines that an interdisciplinary concept comes from, and the knowledge distance between the source discipline and the target discipline.

Measuring semantic gap between user-generated content and product descriptions through compression comparison in e-commerce

The significance of user-generated content as a source for business intelligence and analytics has been on the rise since the inception of electronic commerce platforms and has been solidified in the wake of the pandemic due to the prominence of electronic commerce as a sales channel. The prevailing approach to harnessing unstructured data involves the utilization of Artificial Intelligence; however, there exist simpler alternatives capable of yielding valuable information. This article introduces a methodology grounded in information theory to quantify the semantic disparity between the consumer community and product descriptions.This disparity can result in potential misunderstandings in the dialogue among consumers, and incidental costs in the dialogue between consumers and vendors. One plausible explanation for this disparity is that the terminology employed by consumers may possess different meanings compared to that utilized by product description writers. Our methodology employs large corpora of consumer reviews and product descriptions to quantify this semantic disparity across multiple electronic commerce domains through the implementation of random word exchanges and compression. Furthermore, we utilize neural word embeddings to identify specific words exhibiting the greatest semantic drift between reviews and descriptions, thereby providing lexical examples of these gaps. Our findings indicate that lower levels of lexical-semantic gap are associated with better consumer satisfaction.

Tracking mutational semantics of SARS-CoV-2 genomes

Natural language processing (NLP) algorithms process linguistic data in order to discover the associated word semantics and develop models that can describe or even predict the latent meanings of the data. The applications of NLP become multi-fold while dealing with dynamic or temporally evolving datasets (e.g., historical literature). Biological datasets of genome-sequences are interesting since they are sequential as well as dynamic. Here we describe how SARS-CoV-2 genomes and mutations thereof can be processed using fundamental algorithms in NLP to reveal the characteristics and evolution of the virus. We demonstrate applicability of NLP in not only probing the temporal mutational signatures through dynamic topic modelling, but also in tracing the mutation-associations through tracing of semantic drift in genomic mutation records. Our approach also yields promising results in unfolding the mutational relevance to patient health status, thereby identifying putative signatures linked to known/highly speculated mutations of concern.

Sampling-invariant fully metric learning for few-shot object detection

Few-shot object detection (FSOD) aims to learn models to detect unseen objects with a few annotated exemplars. Despite great success in FSOD, existing metric-based methods heavily rely on class prototypes extracted from limited training data and thus suffer severely from data shift, which can learn inductive bias and produce inconsistent performance across multiple runs. According to statistic law, data shift happens more frequently under low-shot scenarios than that in many-shot scenarios. However, there is insufficient research about utilizing intrinsic and robust properties of limited data for data-stable results instead of blindly achieving high performance on a specific dataset. This inspires us to model intrinsic properties of limited data for building a set of robust class prototypes. To this end, we propose a novel Sampling-Invariant Fully metric learning Network (SIF-Net) that enhances class prototypes in three aspects. Specifically, our model includes three novel modules: (1) a multi-scale context matching (MSCM) module that aggregates more accurate class concepts in a scale-matching manner, (2) a semantic drift correlation (SDC) module that recovers distorted class prototypes through object context, and (3) a fully metric learning (FML) module that encodes both class and spatial priors into class prototypes, which is fundamentally different from previous metric-based approaches. As a result, our SIF-Net is robust to various settings of few-shot data quality. Extensive experiments show that our fully metric-based framework is superior to other metric-based approaches and achieves state-of-the-art performance on PASCAL VOC and MS COCO datasets.

Hybrid Autoregressive Inference for Scalable Multi-Hop Explanation Regeneration

Regenerating natural language explanations in the scientific domain has been proposed as a benchmark to evaluate complex multi-hop and explainable inference. In this context, large language models can achieve state-of-the-art performance when employed as cross-encoder architectures and fine-tuned on human-annotated explanations. However, while much attention has been devoted to the quality of the explanations, the problem of performing inference efficiently is largely under studied. Cross-encoders, in fact, are intrinsically not scalable, possessing limited applicability to real-world scenarios that require inference on massive facts banks. To enable complex multi-hop reasoning at scale, this paper focuses on bi-encoder architectures, investigating the problem of scientific explanation regeneration at the intersection of dense and sparse models. Specifically, we present SCAR (for Scalable Autoregressive Inference), a hybrid framework that iteratively combines a Transformer-based bi-encoder with a sparse model of explanatory power, designed to leverage explicit inference patterns in the explanations. Our experiments demonstrate that the hybrid framework significantly outperforms previous sparse models, achieving performance comparable with that of state-of-the-art cross-encoders while being approx 50 times faster and scalable to corpora of millions of facts. Further analyses on semantic drift and multi-hop question answering reveal that the proposed hybridisation boosts the quality of the most challenging explanations, contributing to improved performance on downstream inference tasks.

Fugitive Archives: Translation, Sea, and History in Indian Ocean Fiction

Abstract This article explores the representation of multilingual Indian Ocean pasts in novels by Amitav Ghosh and Abdulrazak Gurnah, two key contemporary postcolonial writers from the opposite shores of the ocean. It theorizes the historical impulse in the novels as anarchival drift, which refers to the self-conscious mode of rewriting the past that subjects the archive to the instability and fluidity of the sea. Ghosh’s Sea of Poppies (2006) and Gurnah’s Paradise (1994) both tell stories of forced displacements in the nineteenth-century Indian Ocean; both rewrite colonial archives in order to depict cross-cultural interactions, employing various self-reflexive textual strategies that draw attention to the linguistic and archival mediations operating in those encounters. This article examines these textual moments alongside the novels’ archival sources—specifically, nineteenth-century colonial dictionaries and Swahili travelogues—to argue that the self-reflexivity results from the multiplicity of linguistic registers on which these texts operate, making visible the translative processes imbricated in transoceanic historical forces. While both novels appeal to the linguistic aspects of cross-cultural interactions, the article traces the divergent ways in which the semantic drift among languages stage the materiality and historicity of trans-oceanic encounters.