Applying machine learning algorithms to automatically infer relationships between concepts from large-scale collections of documents presents a unique opportunity to investigate at scale how human semantic knowledge is organized, how people use it to make fundamental judgments (“How similar are cats and bears?”), and how these judgments depend on the features that describe concepts (e.g., size, furriness). However, efforts to date have exhibited a substantial discrepancy between algorithm predictions and human empirical judgments. Here, we introduce a novel approach to generating (...) embeddings for this purpose motivated by the idea that semantic context plays a critical role in human judgment. We leverage this idea by constraining the topic or domain from which documents used for generating embeddings are drawn (e.g., referring to the natural world vs. transportation apparatus). Specifically, we trained state-of-the-art machine learning algorithms using contextually-constrained text corpora (domain-specific subsets of Wikipedia articles, 50+ million words each) and showed that this procedure greatly improved predictions of empirical similarity judgments and feature ratings of contextually relevant concepts. Furthermore, we describe a novel, computationally tractable method for improving predictions of contextually-unconstrained embedding models based on dimensionality reduction of their internal representation to a small number of contextually relevant semantic features. By improving the correspondence between predictions derived automatically by machine learning methods using vast amounts of data and more limited, but direct empirical measurements of human judgments, our approach may help leverage the availability of online corpora to better understand the structure of human semantic representations and how people make judgments based on those. (shrink)

Traditional methods for deriving property-based representations of concepts from text have focused on either extracting only a subset of possible relation types, such as hyponymy/hypernymy (e.g., car is-a vehicle) or meronymy/metonymy (e.g., car has wheels), or unspecified relations (e.g., car—petrol). We propose a system for the challenging task of automatic, large-scale acquisition of unconstrained, human-like property norms from large text corpora, and discuss the theoretical implications of such a system. We employ syntactic, semantic, and encyclopedic information to (...) guide our extraction, yielding concept-relation-feature triples (e.g., car be fast, car require petrol, car cause pollution), which approximate property-based conceptual representations. Our novel method extracts candidate triples from parsed corpora (Wikipedia and the British National Corpus) using syntactically and grammatically motivated rules, then reweights triples with a linear combination of their frequency and four statistical metrics. We assess our system output in three ways: lexical comparison with norms derived from human-generated property norm data, direct evaluation by four human judges, and a semantic distance comparison with both WordNet similarity data and human-judged concept similarity ratings. Our system offers a viable and performant method of plausible triple extraction: Our lexical comparison shows comparable performance to the current state-of-the-art, while subsequent evaluations exhibit the human-like character of our generated properties. (shrink)

Cognitive Science, Volume 46, Issue 2, February 2022.

Dans un premier temps, cette contribution fournit une présentation rapide des corpus médiévaux, principalement pour l’anglais et le français, en relation avec des projets d’annotation centrés sur l’annotation syntaxique. La seconde partie traite des aspects spécifiques de l’annotation syntaxique pour les textes d’ancien français et s’intéresse particulièrement aux interactions entre la préannotation automatique et les outils d’annotation manuelle.

Dans un premier temps, cette contribution fournit une présentation rapide des corpus médiévaux, principalement pour l’anglais et le français, en relation avec des projets d’annotation centrés sur l’annotation syntaxique. La seconde partie traite des aspects spécifiques de l’annotation syntaxique pour les textes d’ancien français et s’intéresse particulièrement aux interactions entre la préannotation automatique et les outils d’annotation manuelle.

This book reflects the growing influence of corpus linguistics in a variety of areas such as lexicography, translation studies, genre analysis, and language ...

In this paper, we present an overview of existing parallel corpora for Automatic Text Simplification (ATS) in different languages focusing on the approach adopted for their construction. We make the main distinction between manual and (semi)–automatic approaches in order to investigate in which respect complex and simple texts vary and whether and how the observed modifications may depend on the underlying approach. To this end, we perform a two-level comparison on Italian corpora, since this is the only (...) language, with the exception of English, for which there are large parallel resources derived through the two approaches considered. The first level of comparison accounts for the main types of sentence transformations occurring in the simplification process, the second one examines the results of a linguistic profiling analysis based on Natural Language Processing techniques and carried out on the original and the simple version of the same texts. For both levels of analysis, we chose to focus our discussion mostly on sentence transformations and linguistic characteristics that pertain to the morpho-syntactic and syntactic structure of the sentence. (shrink)

The central aims of this paper are to show how linguistic corpora have been used and can be used in philosophy and to argue that linguistic corpora and corpus analysis should be added to the philosopher’s toolkit of ways to address philosophical questions. A linguistic corpus is a curated collection of texts representing language use that can be queried to answer research questions. Among many other uses, linguistic corpora can help answer questions about the meaning of words (...) and the structure of discourse. Through a discussion of examples, the paper shows that there are many philosophical questions that can be addressed by using a linguistic corpus. However, linguistic corpora need not (and often cannot) replace traditional philosophical methods. Lastly, it argues that the special properties of linguistic corpora, including their independence and the current ease and cheapness of access, make them an indispensable resource for philosophers. (shrink)

This article reports on a corpus-based study of the English translation of Wang Anyi’s award-winning novel, The Song of Everlasting Sorrow [长恨歌, Chang Hen Ge] from the perspective of style. Using the keyword and concordance functions of corpus software packages AntConc and ParaConc, this research focuses on how the translator’s style reveals itself in the target text as well as how the style of the source text is represented in the TT. Findings show that the translators have a (...) preference for contracted negative verb forms and expressions such as “bit” over “little,” making the text more colloquial. In general, the rendering of the ST style tends not to be entirely faithful. A distinction between reader-centered and text-centered keywords is made. While the reader-centered ST keywords are, as expected, largely altered, the translation of text-centered ST keywords is also rewritten, contrary to expectations. Presumably, the translators tended to reduce the ambiguity of the ST, resulting in a more explicit TT. The article argues that the translators chose to rewrite the translation to make it more understandable for the target audience since it concerned a work from a source culture very different from the target culture. (shrink)

The step to e-research in philosophy depends on the availability of high quality, easily and freely accessible corpora in a sustainable format composed from multi-language, multi-script books from different historical periods. Corpora matching these needs are at the moment virtually non-existing. Within @PhilosTei, we have addressed this corpus building problem by developing an open source, web-based, user-friendly workflow from textual images to TEI, based on state-of-the-art open source OCR software, to wit Tesseract, and a multi-language version of TICCL, (...) a powerful OCR post-correction tool. We have demonstrated the utility of the tool by applying it to a multilingual, multi-script corpus of important eighteenth to twentieth-century European philosophical texts. (shrink)

Le présent article se propose de revenir sur le projet national CoMeRe (Communication Médiée par les Réseaux) en insistant sur la complexité du corpus développé. Constitué de quatorze sous-corpus variés, le corpus CoMeRe est un corpus de référence de la communication médiée par les réseaux en français. Quatorze enseignants-chercheurs de huit laboratoires différents se sont impliqués dans le projet et ont été guidés par trois mots clés lors de leurs collaborations : variété, standards et accès ouvert.Le corpus CoMeRe a ainsi (...) été construit sur une hypothèse de variété et contient une gamme étendue d’interactions de la CMR dont nous restituons les différences et les caractéristiques principales (courriels, clavardage, SMS, forums de discussion Internet, blogs, tweets, discussions Wikipédia, interactions provenant de mondes synthétiques). Nous détaillons ensuite comme le corpus CoMeRe a été rendu interopérable : les quatorze sous-corpus ont été standardisés, suivant le modèle de l’espace d’interaction élaboré lors du projet (Chanier & Jin, 2013) d’une part et suivant les propositions de représentation standardisée des corpus de la CMR en TEI (Text Encoding Initiative, 2019) élaborées en lien avec les partenaires européens. Enfin, les collègues tenaient à diffuser le corpus en accès ouvert pour permettre son utilisation par la communauté scientifique. Nous revenons sur les retombées du projet liées à la diffusion du corpus dans la conclusion de notre article. (shrink)

This paper presents a new method for producing a dictionary of subcategorization frames from unlabelled text corpora. It is shown that statistical filtering of the results of a finite state parser running on the output of a stochastic tagger produces high quality results, despite the error rates of the tagger and the parser. Further, it is argued that this method can be used to learn all subcategorization frames, whereas previous methods are not extensible to a general solution to (...) the problem. (shrink)

Due to existence of a huge amount of textual data either on the World Wide Web or in textual databases like PubMed, the development of novel automatic keyphrase extraction methods has emerged as one of the key research problems in recent past. Consequently, a number of machine learning techniques, mostly supervised, have been proposed to extract keyphrases from text documents. But, one of the main bottlenecks that hinders the success of such systems is the requirement of annotated corpora (...) for training purpose. In this paper, we propose the design of a deep text mining system to identify keyphrases in text documents that are either unstructured or semi-structured in nature. The novelty of our system lies in its applicability on a single document, instead of demanding a collection of annotated texts for training, to identify keyphrases embedded within it. The proposed system applies parsing techniques to identify candidate phrases. After mapping the original set of candidate phrases into a low-dimensional space using Singular Value Decomposition, the Markov Clustering technique is applied to cluster related sentences together. Finally, considering each cluster as a document, Latent Dirichlet Allocation is applied to identify feasible keyphrases that are presented to users in non-increasing order of their relevance score values. The efficacy of the proposed system is established through experimentation on datasets from two different domains. On comparative evaluation, we found that the proposed system outperforms KEA and KEA that apply the supervised machine learning approach for automatic keyphrase extraction from text documents. (shrink)

This study investigated cross-cultural differences in individual pride and collective pride between Chinese and Americans using data from text corpora. We found higher absolute frequencies of pride items in the American corpus than in the Chinese corpus. Cross-cultural differences were found for relative frequencies of different types of pride, and some of them depended on the genre of the text corpora. For both blogs and news genres, Americans showed higher frequencies of individual pride items and lower (...) frequencies of relational pride items than did their Chinese counterparts. Cross-cultural differences in national pride, however, depended on the genre: Chinese news genre included more national pride items than its American counterpart, but the opposite was true for the blog genre. We discuss the implications of these results in relation to the existing literature on cultural differences in individual pride and collective pride. (shrink)

Deep Learning has been widely used for tackling challenging natural language processing tasks over the recent years. Similarly, the application of Deep Neural Networks in legal analytics has increased significantly. In this survey, we study the early adaptation of Deep Learning in legal analytics focusing on three main fields; text classification, information extraction, and information retrieval. We focus on the semantic feature representations, a key instrument for the successful application of deep learning in natural language processing. Additionally, we share (...) pre-trained legal word embeddings using the word2vec model over large corpora, comprised legislations from UK, EU, Canada, Australia, USA, and Japan among others. (shrink)

Deep Learning has been widely used for tackling challenging natural language processing tasks over the recent years. Similarly, the application of Deep Neural Networks in legal analytics has increased significantly. In this survey, we study the early adaptation of Deep Learning in legal analytics focusing on three main fields; text classification, information extraction, and information retrieval. We focus on the semantic feature representations, a key instrument for the successful application of deep learning in natural language processing. Additionally, we share (...) pre-trained legal word embeddings using the word2vec model over large corpora, comprised legislations from UK, EU, Canada, Australia, USA, and Japan among others. (shrink)

A significant portion of the world’s text is tagged by readers on social bookmarking websites. Credit attribution is an inherent problem in these corpora because most pages have multiple tags, but the tags do not always apply with equal specificity across the whole document. Solving the credit attribution problem requires associating each word in a document with the most appropriate tags and vice versa. This paper introduces Labeled LDA, a topic model that constrains Latent Dirichlet Allocation by defining (...) a one-to-one correspondence between LDA’s latent topics and user tags. This allows Labeled LDA to directly learn word-tag correspondences. We demonstrate Labeled LDA’s improved expressiveness over traditional LDA with visualizations of a corpus of tagged web pages from del.icio.us. Labeled LDA outperforms SVMs by more than 3 to 1 when extracting tag-specific document snippets. As a multi-label text classifier, our model is competitive with a discriminative baseline on a variety of datasets. (shrink)

Topic model is a well proven tool to investigate the semantic content of textual corpora. Yet corpora sometimes include texts in several languages, making it impossible to apply language-specific computational approaches over their entire content. This is the problem we encountered when setting to analyze a philosophy of science corpus spanning over eight decades and including original articles in Dutch, German and French, on top of a large majority of articles in English. To circumvent this multilingual problem, we (...) use machine-translation tools to bulk translate non-English documents into English. Though largely imperfect, especially syntactically, these translations nevertheless provide correctly translated terms and preserve the semantic proximity of documents with respect to one another. To assess the quality of this translation step, we develop a “semantic topology preservation test” that relies on estimating the extent to which document-to-document distances have been preserved during translation. We then conduct an LDA topic-model analysis over the entire corpus of translated and English original texts, and compare it to a topic-model done over the English original texts only. We thereby identify the specific contribution of the translated texts. These studies reveal a more complete picture of main topics that can found in the philosophy of science literature, especially during the early days of the discipline when numerous articles were published in languages other than English. (shrink)

In this article, we use a mix of computational techniques to identify textual shifts in the Ottoman şeyhülislams’ fetvas between the sixteenth and twentieth centuries. Our analysis, supplemented by a close reading of these texts, indicates that the fetvas underwent multiple forms of transformation, a consequence of the institutional evolution of the şeyhülislam’s fetva office that aimed to speed up and streamline the production of the fetvas: over time, the texts appropriated a more uniform character and came to contain shorter (...) responses. In the compositions of the questions, we identified many “trigger terms” that facilitated reflexive responses independent of the fetvas’ jurisprudential contexts, a tendency that became stronger after the second half of the seventeenth century. In addition, we propose in the article a methodology that measures the relative strengths of textual and conceptual links among the fetva corpora of various Ottoman şeyhülislams. This analysis informs us about possible paths of long-term evolution of this genre of jurisprudential documents. (shrink)

Nous examinons ici les schémas lexico-grammaticaux (LG) associés à trois verbes supports (« effectuer », « réaliser », « procéder ») dans deux corpus juridiques comparables (un corpus de directives européennes rédigées en français et leur transposition en France dans les textes de lois). A cette fin, nous utilisons Le Trameur, un logiciel de textométrie qui permet une comparaison systématique des données textuelles en mobilisant plusieurs méthodes statistiques exploratoires : analyse des fréquences par partie, spécificités, segments répétés, co-occurrences, poly-cooccurrences, etc. (...) Le recours à ce type d’outils permet au linguiste d’élaborer des unités de base de son système analytique en suivant un protocole de recherche précis, qui tient compte des indices quantitatifs. Finalement, des distinctions lexico-grammaticales fines entre des lexèmes dont le sémantisme serait trop proche pour une comparaison hors contexte, sont visualisées de façon synthétique au fil du texte, tout en permettant de différentier certaines constructions spécifiques du français « eurolectal » et les éléments du discours technocratique « national ». (shrink)

The aim of this paper is to explore some of the possibilities, advantages and difficulties of corpus-based analyses of semantic and pragmatic aspects of language in one particular field, namely the linguistic expression of emotion concepts. For this purpose, a methodological procedure is proposed and an exemplary analysis of the emotion concept “fear” in English is performed. The procedure combines Kövecses' lexical approach and Stefanowitsch's metaphorical pattern analysis with additional concepts from corpus linguistics such as semantic preference and semantic prosody. (...) The results of the study show that such a corpus-based analysis of emotion words offers several advantages. Firstly, by exploring the surroundings of the search word in a vast amount of text, we are not only able to find evidence of conceptual metaphor and metonymy that structure the emotion concept and of related emotion concepts, but also we can enrich the description of the emotion concept with information from a series of dimensions and add a pragmatic viewpoint by revealing an explicit or implicit evaluation of the emotion. The second advantage offered by a corpus-based approach lies in the possibility of quantifying results, i.e., comparing the frequency, productivity and creative use of individual metaphors and metonymies, which is especially interesting in view of contrastive studies. (shrink)

Un des objectifs des niveaux élémentaire et indépendant du Cadre européen commun de référence pour les langues (CECRL) est d’amener l’apprenant — alors qu’il n’est pas encore un utilisateur autonome de la langue — à être capable d’interagir dans les situations authentiques de la vie quotidienne. Or, la plupart des matériaux pédagogiques consacrés à ces niveaux ne présentent qu’un échantillonnage partiel et souvent incomplet du langage authentique. Dans cet article, nous proposons une méthodologie dédiée, d’une part, à la construction de (...) corpus thématiques à des fins pédagogiques aux niveaux A1+ à B1 du CECRL et, d’autre part, à l’identification du vocabulaire-clé associé.Notre approche repose sur une sélection thématique de textes allemands selon les sujets définis par le CECRL afin de créer (1) un corpus ciblé et (2) de taille relativement faible, (3) regroupant des textes (quasi-)authentiques, adaptés aux niveaux des apprenants. En accord avec les résultats issus de la linguistique de corpus, soulignant la place centrale des unités multi-lexicales (UML) dans la construction du vocabulaire, le vocabulaire-clé tiré de cette collection consiste en UML plutôt qu’en mots individuels. Le procédé de sélection est illustré à travers l’exemple concret d’un sous-thème de la catégorie « Essen und Trinken » (Nourriture) du CECRL formé à partir d’un corpus de critiques de restaurants. Nous concluons par le constat que cette approche empirique peut mener à un changement dans la perception de la nature du vocabulaire-clé susceptible d’être présenté aux apprenants aux niveaux de compétences linguistiques inférieurs. (shrink)

Social scientists and computer scientist are divided by small differences in perspective and not by any significant disciplinary divide. In the field of text analysis, several such differences are noted: social scientists often use unsupervised models to explore corpora, whereas many computer scientists employ supervised models to train data; social scientists hold to more conventional causal notions than do most computer scientists, and often favor intense exploitation of existing algorithms, whereas computer scientists focus more on developing new models; (...) and computer scientists tend to trust human judgment more than social scientists do. These differences have implications that potentially can improve the practice of social science. (shrink)

A lot of effort is currently made to provide methods to analyze and understand deep neural network impressive performances for tasks such as image or text classification. These methods are mainly based on visualizing the important input features taken into account by the network to build a decision. However these techniques, let us cite LIME, SHAP, Grad-CAM, or TDS, require extra effort to interpret the visualization with respect to expert knowledge. In this paper, we propose a novel approach to (...) inspect the hidden layers of a fitted CNN in order to extract interpretable linguistic objects from texts exploiting classification process. In particular, we detail a weighted extension of the Text Deconvolution Saliency (wTDS) measure which can be used to highlight the relevant features used by the CNN to perform the classification task. We empirically demonstrate the efficiency of our approach on corpora from two different languages: English and French. On all datasets, wTDS automatically encodes complex linguistic objects based on co-occurrences and possibly on grammatical and syntax analysis. (shrink)

Over the past two decades, researchers have made great advances in the area of computational methods for extracting meaning from text. This research has to a large extent been spurred by the development of latent semantic analysis (LSA), a method for extracting and representing the meaning of words using statistical computations applied to large corpora of text. Since the advent of LSA, researchers have developed and tested alternative statistical methods designed to detect and analyze meaning in (...) class='Hi'>text corpora. This research exemplifies how statistical models of semantics play an important role in our understanding of cognition and contribute to the field of cognitive science. Importantly, these models afford large-scale representations of human knowledge and allow researchers to explore various questions regarding knowledge, discourse processing, text comprehension, and language. This topic includes the latest progress by the leading researchers in the endeavor to go beyond LSA. (shrink)

Our methods of research, duly elaborated hereafter, would benefit from being applied to the realm of the East. For that matter, the examination of Syriac, Armenian, Coptic or Arabic manuscripts does not differ in the least from that of a Greek or Latin manuscript. The rules developed by classical philologists are just as valid for the study of the Maxims of Phtahhotep and the Precepts of Kagemeni…Alphonse Dain (1975), Les Manuscrits (Paris, Les Belles Lettres)One of the objects of a comparative (...) history of the scholarly practices in the West, the Far East, the Arab world and classical antiquity is the role of libraries in the constitution, transmission and transformations of the great textual corpora. To the institutional and social aspects of the politics of social memory should be added the range of intellectual and technical practices which have as their aim the conservation of this inheritance: material conservation and transmission - copying texts, restoring them, transferring them from one medium, one script and one language to another - but also work on the literal level and the meaning, the content and the form. (shrink)

The paper explores the role of competing notions of what does it mean to have a testament of the law of the past in Christian and Rabbinic corpora of text and thought. The argument probes and renegotiates the complex relationships of the Christian suspension of Old Testament by the New Testament and the Rabbinic suspension of new testament in the two Talmudim. It consequently draws implications of that analysis for understanding the relationships of the two Talmudim to the (...) tradition of hermeneutics of texts, as influenced as the latter has been by theological and literary approaches of various Christian theologies of the two Testaments. As a part of that analysis the articles justifies the task of advancing and providing a critique of political theology and political philology as modes of thought and investigation. That provides a way to ask anew the question about relationships between theology, literary theory, and political thought. (shrink)

This paper proposes several recurrent neural network-based models for recognizing requisite and effectuation parts in Legal Texts. Firstly, we propose a modification of BiLSTM-CRF model that allows the use of external features to improve the performance of deep learning models in case large annotated corpora are not available. However, this model can only recognize RE parts which are not overlapped. Secondly, we propose two approaches for recognizing overlapping RE parts including the cascading approach which uses the sequence of BiLSTM-CRF (...) models and the unified model approach with the multilayer BiLSTM-CRF model and the multilayer BiLSTM-MLP-CRF model. Experimental results on two Japan law RRE datasets demonstrated advantages of our proposed models. For the Japanese National Pension Law dataset, our approaches obtained an \ score of 93.27% and exhibited a significant improvement compared to previous approaches. For the Japan Civil Code RRE dataset which is written in English, our approaches produced an \ score of 78.24% in recognizing RE parts that exhibited a significant improvement over strong baselines. In addition, using external features and in-domain pre-trained word embeddings also improved the performance of RRE systems. (shrink)

By the time they reach early adulthood, English speakers are familiar with the meaning of thousands of words. In the last decades, computational simulations known as distributional semantic models have demonstrated that it is possible to induce word meaning representations solely from word co-occurrence statistics extracted from a large amount of text. However, while these models learn in batch mode from large corpora, human word learning proceeds incrementally after minimal exposure to new words. In this study, we run (...) a set of experiments investigating whether minimal distributional evidence from very short passages suffices to trigger successful word learning in subjects, testing their linguistic and visual intuitions about the concepts associated with new words. After confirming that subjects are indeed very efficient distributional learners even from small amounts of evidence, we test a DSM on the same multimodal task, finding that it behaves in a remarkable human-like way. We conclude that DSMs provide a convincing computational account of word learning even at the early stages in which a word is first encountered, and the way they build meaning representations can offer new insights into human language acquisition. (shrink)

Cet article propose d'interroger la notion de contextualisation à partir d'une perspective d'analyse textuelle des discours. La contextualisation est ici définie comme le processus par lequel le chercheur tente d’établir la pertinence d’une mise en relation entre un texte et un autre (ou plusieurs autres) et de leur regroupement au sein d’un corpus. Trois critères (générique, auctorial et thématique) sont identifiés, autorisant les passages de texte à texte(s) et dessinant ce que nous suggérons d'appeler des sphères de contextualisation. Finalement, une (...) hiérarchisation et des articulations possibles entre ces trois sphères sont envisagées. (shrink)

Cet article propose d'interroger la notion de contextualisation à partir d'une perspective d'analyse textuelle des discours. La contextualisation est ici définie comme le processus par lequel le chercheur tente d’établir la pertinence d’une mise en relation entre un texte et un autre (ou plusieurs autres) et de leur regroupement au sein d’un corpus. Trois critères (générique, auctorial et thématique) sont identifiés, autorisant les passages de texte à texte(s) et dessinant ce que nous suggérons d'appeler des sphères de contextualisation. Finalement, une (...) hiérarchisation et des articulations possibles entre ces trois sphères sont envisagées. (shrink)

The Corpus of English Philosophy Texts (CEPhiT) is part of the Coruña Corpus of English Scientific Writing (CC) following CETA (Corpus of English Texts on Astronomy). In accordance with the rest of the Corpus, CEPhiT has been compiled for the description of philosophical texts written in English in the eighteenth and nineteenth centuries. The sampling method employed requires the collection of extracts of 10,000 words. This method has been followed in CETA and CEPhiT, with samples from 40 different authors in (...) the latter, both from Europe and North America. Text selection is based on some extralinguistic criteria, such as year of publication, sex, geographical provenance and text-types/genres. The corpus contains samples belonging to six different genre categories. This taxonomy, as well as some other extralinguistic information, can be used to search the corpus. CEPhiT, like CETA, is released on CD-Rom with Coruña Corpus Tool (CCT), purpose-designed software by IrLab. (shrink)

The opinion according to which Thomas Aquinas was the original owner of ms. Vat. lat. 718 has proved wrong and cannot serve anymore for dating the codex before Thomas’ departure to Italy . The Parisian tradition of De progressu animalium and De motu animalium integrated Moerbeke’s two translations not only into the Aristotelian Corpus recentius of zoological treatises, but also into the corpora recentiora of the parua. It is probable that De historia animalium was augmented with Book X in (...) a secondary stage of its circulation. As far as the university tradition of De motibus animalium et aliorum paruorum is concerned, not all questions regarding three successive exemplaria, and more precisely those of the exact contents, composition and number of peciae are definitively resolvable for the moment. R.-A. Gauthier’s position regarding the composition of the exemplar of the psychological parua in Np3 must be corrected after new examination of the manuscript witnesses . Several Italian witnesses do not depend on the university text, but regularly attest to the existence of independent traditions. Moerbeke’s Greek codex used in his first version of the Latin translation has some similarities with the Greek codex used by the Anonymous whose translation became the basis of Albert the Great’s De principiis motus processivi. This latter Greek codex is not preserved, but it had some striking similarities with ms. Firenze Laur. LXXXVII.21, which originated in the monastery of Hagios Nikalaos in Casole near Otranto. Moerbeke’s Greek model was certainly not identical with that used by the Anonymous and was, indeed, obviously better. (shrink)

Scribal copying is investigated as a test case for the memetic and epidemiological models for explaining the distribution of cultural items. We may hypothesize that the incidence of errors could be low enough to allow two conditions for neo-Darwinian explanation to be fulfilled: first, that there be a rather reliable mechanism for heredity, and second that occasional mutations might produce a version more likely to survive and be propagated than the exemplar. Scriptorial conventions are reviewed. Textual criticism is investigated. Finally, (...) some attention is given to the psychology of language perception and production. In conclusion, it is argued that the memetic model for cultural transmission is not generally fecund while the epidemiological model is. (shrink)

Corpus-based word frequencies are one of the most important predictors in language processing tasks. Frequencies based on conversational corpora are shown to better capture the variance in lexical decision tasks compared to traditional corpora. In this study, we show that frequencies computed from social media are currently the best frequency-based estimators of lexical decision reaction times. The results are robust and are still substantial when we control for corpus size.