Papers reviewed and determined not to be word norm studies. Use the flag icon to report errors or suggest re-inclusion.
18265 papers
Las gramaticas independientes del contexto son una forma utilizada habitualmente para representar la estructura sintactica de las oraciones. Muchas tareas dedicadas al procesamiento del lenguaje natural necesitan de dicha estructura para poder interpretar cada frase. Sin embargo, el problema de la ambiguedad sintactica es muy comun (sobre todo en frases con mas de 15 palabras) y dificulta enormemente la labor. Algunos autores establecen que la mayoria de ambiguedades sintacticas se pueden resolver sin utilizar informacion semantica alguna, esto es, solo seleccionando el analisis sintactico mas probable de entre todos los candidatos. Esto establece las bases de una familia de tecnicas que utilizan las probabilidades para decidir cual es el analisis sintactico que mejor se adapta a cada frase. Las probabilidades de cada estructura se estiman a partir de corpus de frases analizadas sintacticamente (treebanks). El Penn Tree-bank es un ejemplo de este tipo de corpus. La tecnica mas conocida que construye gramaticas independientes del contexto probabilisticas (GICP) a partir de treebanks es aquella en la que las reglas se extraen directamente de los arboles de analisis sintacticos. Las probabilidades se estiman contando el numero de veces que cada produccion aparece. Como resultado se obtienen las denominadas treebank grammars. Este es el esquema mas sencillo y no esta libre de problemas, por una parte, porque las treebank grammars sobre-generalizan en exceso lo que evidencia la necesidad del uso de mecanismos adicionales para relajar la independencia de la GICP y, por otra, porque los analisis sintacticos resultan ser extremadamente planos y por tanto con escasa utilidad semantica. Con este espiritu se ha trabajado en esta tesis doctoral con una generalizacion de los clasicos k-gramas pero aplicados a arboles en lugar de cadenas. Las GICP obtenidas consisten en reglas que incluyen informacion contextual acerca de cuando pueden ser aplicadas. Los resultados han sido fructiferos: se ha reducido drasticamente la ambiguedad de las gramaticas y por tanto tambien el tiempo de proceso de los analizadores sintacticos cuyas respuestas resultan ser, ademas, mas refinadas y mejor estructuradas de manera que son provechosas para otros fines linguisticos. En contrapartida y como cabia esperar, se ha producido el efecto contrario a la sobre-generalizacion: la asignacion de probabilidad nula, y por tanto rechazo, de ciertas frases sintacticamente correctas. Este es el hecho que ha motivado la investigacion sobre mecanismos de suavizado de las funciones de probabilidad que se adaptaran de forma eficiente a las nuevas gramaticas introducidas. Las aportaciones fundamentales son: 1.Una descripcion de como los lenguajes k-testables probabilisticos de arboles se pueden aproximar a cualquier lenguaje de arboles racional estocastico. 2. Un metodo para aprender eficientes gramaticas independientes del contexto probabilisticas a partir de lenguajes de arboles. 3. Un analisis de distintas tecnicas de suavizado de las funciones de probabilidad. 4. Un metodo de construccion de gramaticas k-testables con suavizado implicito. 5. Experimentos con fructiferos resultados: se ha reducido drasticamente la ambiguedad de las gramaticas y por tanto tambien el tiempo de proceso de los analizadores sintacticos cuyas respuestas resultan ser, ademas, mas refinadas y mejor estructuradas de manera que son provechosas para otros fines linguisticos.
Afro-Linguistics is a branch of linguistics that focuses on languages spoken in Africa and their structural characteristics. The study employs a combination of qualitative interviews with language experts and quantitative analysis of linguistic databases to assess the diversity and complexity of Ghanaian languages. A notable finding is that over 40% of Ghana's indigenous languages exhibit complex grammatical structures, indicating the need for comprehensive policy frameworks to support their continued use and study. The research underscores the importance of integrating Afro-Linguistic insights into language preservation policies in Ghana, particularly focusing on enhancing educational resources and community engagement initiatives. Recommendations include the development of a national register of endangered languages, allocation of funds for linguistic research centers, and the integration of Afro-Linguistics principles into primary school curricula.
This paper explores ways to detect errors in aligned corpora, using very little technology. In the first method, applicableto any aligned corpus, we consider alignment as a string-to-string mapping...
Proceedings of the Ninth International Workshop \non Treebanks and Linguistic Theories. \nEditors: Markus Dickinson, Kaili Müürisep and Marco Passarotti. \nNEALT Proceedings Series, Vol. 9 (2010), 223-232. \n© 2010 The editors and contributors. \nPublished by \nNorthern European Association for Language \nTechnology (NEALT) \nhttp://omilia.uio.no/nealt. \nElectronically published at \nTartu University Library (Estonia) \nhttp://hdl.handle.net/10062/15891.
Obesity prevalence in the U.S. has increased during the last three decades with major impact on public health. Screening for obesity in a population with unknown weight status can be time- and resource-consuming, but the information is valuable for prioritizing and allocating scarce resources. The challenge remains to properly assess obesity with the available methods. Body Image Rating Scales (BIRS) have initially been developed to assess body image disturbances, but also seem useful as an alternative method in assessing obesity prevalence. Several different BIRS exists. In this project I reviewed the literature that exists regarding the use of BIRS, and its advantages and limitations for the assessment of obesity status with regards to BMI. The result yielded nine publications that examined eight different scales and their correlation with BMI, ranging from r=.59 for self-reported BMI to r=.94 for measured BMI. One concern is the lack of standardization of this method to assess obesity, given the range of different scales. While many methods for obesity assessment are available, the simplicity, ease of use and cost-effectiveness of BIRS make it very appealing. BIRS remain a potentially attractive option to assess the weight status of a large population with minimal requirements in assets and time, especially in situations where measuring instruments are not available, or when height or weight could not be recalled.
In this paper, we present a novel approach to enhance hierarchical phrase-based machine translation systems with linguistically motivated syntactic features. Rather than directly using treebank categories as in previous studies, we learn a set of linguistically-guided latent syntactic categories automatically from a source-side parsed, word-aligned parallel corpus, based on the hierarchical structure among phrase pairs as well as the syntactic structure of the source side. In our model, each X nonterminal in a SCFG rule is decorated with a real-valued feature vector computed based on its distribution of latent syntactic categories. These feature vectors are utilized at decoding time to measure the similarity between the syntactic analysis of the source side and the syntax of the SCFG rules that are applied to derive translations. Our approach maintains the advantages of hierarchical phrase-based translation systems while at the same time naturally incorporates soft syntactic constraints.
We describe a method for the automatic extraction of a Stochastic Lexicalized Tree Insertion Grammar from a linguistically rich HPSG Treebank. The extraction method is strongly guided by HPSG–based head and argument decomposition rules. The tree anchors correspond to lexical labels encoding fine–grained information. The approach has been tested with a German corpus achieving a labeled recall of 77.33% and labeled precision of 78.27%, which is competitive to recent results reported for German parsing using the Negra Treebank.
This paper describes the development of a hybrid tool for a semi-automated process for validation of treebank annotation at various levels. The tool is developed for error detection at the part-of-speech, chunk and dependency levels of a Hindi treebank, currently under development. The tool aims to identify as many errors as possible at these levels to achieve consistency in the task of annotation. Consistency in treebank annotation is a must for making data as error-free as possible and for providing quality assurance. The tool is aimed at ensuring consistency and to make manual validation cost effective. We discuss a rule based and a hybrid approach (statistical methods combined with rule-based methods) by which a high-recall system can be developed and used to identify errors in the treebank. We report some results of using the tool on a sample of data extracted from the Hindi treebank. We also argue how the tool can prove useful in improving the annotation guidelines which would in turn, better the quality of annotation in subsequent iterations. 1.
Two of the main corpora available for training discourse relation classifiers are the RST Discourse Treebank (RST-DT) and the Penn Discourse Treebank (PDTB), which are both based on the Wall Street Journal corpus. Most recent work using discourse relation classifiers have employed fully-supervised methods on these corpora. However, certain discourse relations have little labeled data, causing low classification performance for their associated classes. In this paper, we attempt to tackle this problem by employing a semi-supervised method for discourse relation classification. The proposed method is based on the analysis of feature cooccurrences in unlabeled data. This information is then used as a basis to extend the feature vectors during training. The proposed method is evaluated on both RST-DT and PDTB, where it significantly outperformed baseline classifiers. We believe that the proposed method is a first step towards improving classification performance, particularly for discourse relations lacking annotated data.
Abstract We investigate the use of instance-based ranking methods for surface realization in natural language generation. Our approach to instance-based natural language generation (IBNLG) employs two components: a rule system that ‘overgenerates’ a number of realization candidates from a meaning representation and an instance-based ranker that scores the candidates according to their similarity to examples taken from a training corpus. We develop an efficient search technique for identifying the optimal candidate based on a novel extension of the A * algorithm. The rule system is produced automatically from a semantically annotated fragment of the Penn Treebank II containing management succession texts. We detail the annotation scheme and grammar induction algorithm and evaluate the efficiency and output of the generator. We also discuss issues such as input coverage (completeness) and fluency that are relevant to surface generation in general.
In this paper, we present an on-going project aiming at extending the WordNet lexical database by encoding common sense featural knowledge elicited from language speakers. Such extension of WordNet is required in the framework of the STaRS.sys project, which has the goal of building tools for supporting the speech therapist during the preparation of exercises to be submitted to aphasic patients for rehabilitation purposes. We review some preliminary results and illustrate what extensions of the existing WordNet model are needed to accommodate for the encoding of commonsense (featural) knowledge. 1
The use of images of older people in the British advertising media has been under-researched to date. Further, previous research in any country has tended to examine such images from an a priori framework of general impressions and stereotypes of older people. This study addresses these issues with British consumers' (n = 106) impressions, trait ascriptions, and similarity-between-images ratings of a representative sample of U.K. magazine advertisements featuring older characters. After a series of sorting task laboratory sessions, multidimensional scaling and hierarchical cluster analyses revealed four clearly defined groups representing types of portrayals. These types emerged from the advertisements and from the views of the consumers themselves. These emergent groupings are: (1) Frail and Vulnerable, (2) Happy and Affluent, (3) Mentors, (4) Active and Leisure-oriented older adults. These groupings seem to be a logical context-appropriate derivation from previous findings on generally held stereotypes of older persons. It is argued that the groupings have the potential to contribute to a reliable typology of advertising portrayals of older people, with potential heuristic leverage in social scientific research of intergenerational communication, lifespan concerns, and the aging process.
This paper presents an efficient approach to use small syntactic category for constructing large tree with the assist of linguistic knowledge for parsing Thai language. VB-EM algorithm is exploited to adjust parameters of any trees for selecting the best tree. Our best result is 70.62% accuracy of bracketing recovery.
The current study presents a conversion and unification of the Penn Discourse TreeBank 2.0 under the XML format. The converted corpus allows for a simultaneous search for syntactically specified discourse information on the basis of the XQuery standard.
This paper deals with the issue of linguistic database summaries understood as linguistically quantified propositions. We show basic capabilities of Quantirius, our interactive system supporting the mining and the assessment of linguistic summaries in a database. The mining of linguistic summaries is realized by means of the concept of a protoform introduced by Zadeh. The assessment of validity degrees of summaries is done via Zadeh's fuzzy logic based calculus, extended by the use of cardinality patterns and triangular norms. The main part of the paper presents an idea of a further processing of the set of generated summaries. The proposed algorithm is composed of a reduction mechanism of summaries based on linguistic terms inclusion and a reduction of summaries by means of the overlapping unimodal linguistic terms. We then present an idea of the approval threshold computing for the truth degrees of summaries. Finally, we show that summaries generated by different protoforms can be joined in a master-detail-like relation, making an interesting structure of information contained in the database.
We carry out two studies on affective state modeling for communication settings that involve unilateral intent on the part of one participant (the evoker) to shift the affective state of another participant (the experiencer). The first investigates viewer response in a narrative setting using a corpus of docu-mentaries annotated with viewer-reported narrative peaks. The second investigates affective triggers in a conversational set-ting using a corpus of recorded interactions, annotated with continuous affective ratings, between a human interlocutor and an emotionally colored agent. In each case, we build a “one-sided ” model using indicators derived from the speech of one participant. Our classification experiments confirm the viabil-ity of our models and provide insight into useful features. Index Terms: affect, speech recognition, audio analysis, natural language communication
This article discusses the treatment of collocations in the context of along-term project on the development of multilingual NLP tools. Besides“classical” two-word collocations, we will focus on the case of complexcollocations (3 words or more) for which a recursive design is presented in theform of collocation of collocations. Although comparatively less numerous thantwo-word collocations, the complex collocations pose important challenges forNLP. The article discusses how these collocations are retrieved from corpora,inserted and stored in a lexical database, how the parser uses such knowledgeand what are the advantages offered by a recursive approach to complexcollocations.
We propose a unified model of syntax and discourse in which text structure is viewed as a tree structure augmented with anaphoric relations and other secondary relations. We describe how the model accounts for discourse connectives and the syntax-discourse-semantics interface. Our model is dependency-based, ie, words are the basic building blocks in our analyses. The analyses have been applied cross-linguistically in the Copenhagen Dependency Treebanks, a set of parallel treebanks for Danish, English, German, Italian, and Spanish which are currently being annotated with respect to discourse, anaphora, syntax, morphology, and translational equivalence.
FrameSQL is a web-based application which the author (Sato, 2003; Sato 2008) created originally for searching the Berkeley FrameNet lexical database. FrameSQL now can handle the Japanese lexical database built by the Japanese FrameNet project (JFN) of Keio University in Japan. FrameSQL can search and view the JFN data released in March of 2009 on a standard web browser. Users do not need to install any additional software tools to use FrameSQL, nor do they even need to download the JFN data to their local computer, because FrameSQL accesses the database of the server computer, and executes searches. FrameSQL not only shows a clear view of the headword’s grammar and combinatorial properties of the database, but also relates a Japanese word with its counterparts in English. FrameSQL puts together the Japanese and English lexical databases, and the user can access them seamlessly, as if they were a unified database. Mutual hyperlinks among these databases and the bilingual search mode make it easy to compare semantic structures of corresponding lexical units between these languages, and it could be useful for building multilingual lexical resources.
Finding a class of structures that is rich enough for adequate linguistic representation yet restricted enough for efficient computational processing is an important problem for dependency parsing. In this paper, we present a transition system for 2-planar dependency trees – trees that can be decomposed into at most two planar graphs – and show that it can be used to implement a classifier-based parser that runs in linear time and outperforms a stateof-the-art transition-based parser on four data sets from the CoNLL-X shared task. In addition, we present an efficient method for determining whether an arbitrary tree is 2-planar and show that 99 % or more of the trees in existing treebanks are 2-planar. 1
This paper proposes a method of correcting annotation errors in a treebank. By using a synchronous grammar, the method transforms parse trees containing annotation errors into the ones whose errors are corrected. The synchronous grammar is automatically induced from the treebank. We report an experimental result of applying our method to the Penn Treebank. The result demonstrates that our method corrects syntactic annotation errors with high precision. 1
With the availability of limited electronic resources, development of a syntactic parser for all types of sentence forms is a challenging and demanding task for any natural language. This paper presents the development of Penn Treebank based statistical syntactic parsers for two South Dravidian languages namely Kannada and Malayalam. Syntactic parsing is the task of recognizing a sentence and assigning a syntactic structure to it. A syntactic parser is an essential tool used for various natural language processing (NLP) applications and natural language understanding. The well known grammar formalism called Penn Treebank structure was used to create the corpus for proposed statistical syntactic parsers. Both the parsing systems were trained using Treebank based corpus consists of 1,000 Kannada and Malayalam sentences that were carefully constructed. The developed corpus has been already annotated with correct segmentation and Part-Of-Speech (POS) information. We have used our own POS tagger generator for assigning proper tags to each and every word in the training and test sentences. The proposed syntactic parser was implemented using supervised machine learning and probabilistic context free grammars (PCFG) approaches. Training, testing and evaluations were done by support vector method (SVM) algorithms. From the experiment we found that the performance of our systems are significantly well and achieves a very competitive accuracy.
A variety of query systems have been developed for interrogating parsed corpora, or treebanks. With the arrival of efficient, widecoverage parsers, it is feasible to create very large databases of trees. However, existing approaches that use in-memory search, or relational or XML database technologies, do not scale up. We describe a method for storage, indexing, and query of treebanks that uses an information retrieval engine. Several experiments with a large treebank demonstrate excellent scaling characteristics for a wide range of query types. This work facilitates the curation of much larger treebanks, and enables them to be used effectively in a variety of scientific and engineering tasks. 1
This paper reports on a treebanking project where eight different modern Chinese translations of the Bible are syntactically analyzed. The trees are created through dynamic treebanking which uses a parser to produce the trees. The trees have been going through manual checking, but correc-tions are made not by editing the tree files but by re-generating the trees with an updated grammar and dictionary. The accuracy of the treebank is high due to the fact that the grammar and dictionary are optimized for this specif-ic domain. The tree structures essen-tially follow the guidelines of the Penn Chinese Treebank. The total number of characters covered by the treebank is 7,872,420 characters. The data has been used in Bible translation and Bi-ble search. It should also prove useful in the computational study of the Chi-nese language in general. 1
In this paper we explore the potential and limitations of a concept of building a bilingual valency lexicon based on the alignment of nodes\nin a parallel treebank. Our aim is to build an electronic Czech<->English Valency Lexicon by collecting equivalences from bilingual\ntreebank data and storing them in two already existing electronic valency lexicons, PDT-VALLEX and Engvallex. For this task a special\nannotation interface has been built upon the TrEd editor, allowing quick and easy collecting of frame equivalences in either of the source\nlexicons. The issues questioning the annotation practice encountered during the first months of annotation include limitations of technical\ncharacter, theory-dependent limitations and limitations concerning the achievable degree of quality of human annotation. The issues of\nspecial interest for both linguists and MT specialists involved in the project include linguistically motivated non-balance between the\nframe equivalents, either in number or in type of valency participants. The first phases of annotation so far attest the assumption that\nthere is a unique correspondence between the functors of the translation-equivalent frames. Also, hardly any linguistically significant\nnon-balance between the frames has been found, which is partly promising considering the linguistic theory used and partly caused by\nlittle stylistic variety of the annotated corpus texts.
Fundamentally, the goals of education are not unlike the goals of medical therapy; that is, to facilitate a desired change in an individual. In medicine, the prerequisite to the selection of any therapeutic regime is the diagnostic process — the identification of the antecedent states of an individual which must be taken into account in the attainment of the intended state. The educational analogue to the medical diagnostic process is the process of identifying the antecedent knowledge, skills, values or attitudes possessed by students entering a course which may influence the process of attaining, or the attainment of, the course goals. The educational analogue of the therapeutic regime are the instructional strategies which take the antecedent conditions into account and which are directed at the fulfillment of course goals. In education however, feasible methods have hot been identified for taking these antecedent conditions into account. In selecting teaching strategies in most classroom situations, it is not practical to take these conditions into consideration on an individual basis. Nor is it useful to consider class averages on these variables, since students vary so widely in terms of them. This study, working in the context of science education, and dealing with affective variables, developed a procedure for providing knowledge of antecedent affective variables in a form permitting their effective utilization in the process of selecting instructional strategies. More specifically, the purpose of this study was to develop a theoretically based and methodologically sound systematic procedure (generically "the Procedure") for (1) identifying, describing, and reporting the degree of pro-ness or con-ness of affective antecedents deemed to be important to science instruction and (2) identifying teaching strategies which take these antecedent conditions into account and which are directed toward science teaching outcomes in the affective domain. The general approach taken by the Procedure is to identify and describe subgroups of students within a class in terms of similar sets of antecedent affective responses to objects which reflect pro-ness or con-ness toward the affective ratings inherent in the affective goals of a course. Instructional strategies for these subgroups can then be selected or provide a rational basis for changing those antecedent ratings which are most incongruent with the desired affective ratings reflected in the affective goals. The affective goals are identified within a clear and accurate statement of the rationale for a course. Measurements of the degree of students' pro-ness or con-ness on the affective responses of concern are obtained through the use of the Semantic Differential technique. The Q-analysis technique, a technique for categorizing people, is employed to identify the subgroups of students. The educational value of the Procedure rests upon its ability to meet an important educational need in a practical way — specifically its ability to provide a clear description of affective antecedents in a form permitting their effective utilization in the process of identifying teaching strategies directed toward the fulfillment of affective goals. In this study, the results of the application of the Procedure to an introductory university physics course supported the general effectiveness of the contribution of each component of the Procedure in meeting this need. There is concern however (1) that additional data need to be gathered supporting the validity of the Procedure, and (2) that the time and monetary demands associated with the Q-analysis and Semantic Differential techniques might limit the feasibility of the Procedure in some educational settings. Recommendations and guidelines regarding future applications of the Procedure are provided, including recommendations regarding validity studies and the use of less costly alternatives to the SD and Q-analysis components.
Exclusive experience with standardized languages produces an expectation that speech community members will assess competing variants by their supposed correctness. This does not happen in the persistently divergent but socially neutral variant use of Embo Gaelic speakers, including close kin and age mates. Communities elsewhere show both variation akin to Embo's personal‐pattern variation and similar social‐organization and social‐setting features: small population size, isolation, use of a minority language unrelated/distantly related to the dominant language, weak/absent extra‐community linguistic norms, social homogeneity with high interaction density and multiplex social roles, absence of linguistic accommodation. The ideologies of linguists from dominant‐language societies hamper recognition of different linguistic ideologies among other peoples. The importance of personal‐pattern variation to individual voice in homogeneous communities is discussed and exemplified. Ongoing transmission of individual variant‐use patterns is seen to have been in keeping with high‐variation language input to children and with diffuse child‐care patterns and peer‐group structure.
Due to idiosyncrasies in their syntax, semantics or frequency, Multiword Expressions (MWEs) have received special attention from the NLP community, as the methods and techniques developed for the treatment of simplex words are not necessarily suitable for them. This is certainly the case for the automatic acquisition of MWEs from corpora. A lot of effort has been directed to the task of automatically identifying them, with considerable success. In this paper, we propose an approach for the identification of MWEs in a multilingual context, as a by-product of a word alignment process, that not only deals with the identification of possible MWE candidates, but also associates some multiword expressions with semantics. The results obtained indicate the feasibility and low costs in terms of tools and resources demanded by this approach, which could, for example, facilitate and speed up lexicographic work.
It was evaluated the effect of environmental alteration in affect, perceived exertion and in physiological parameters during treadmill running in athletes. Eighteen handball players participated in four running experimental sessions (baseline, fragmented overrated and unknown) for 20 minutes. It was analyzed: affect, rating of perceived exertion, heart rate and cortisol. Information about the running duration was manipulated. It was observed a statistically significant increase in affect during the treadmill running for the “fragmented” and “overrated ” sessions, as well as in salivary cortisol level at the 11th minute of running. The control of athletes’ affective behavior seems to relieve physiological damage when facing unknown opponents and uncertain game conditions.
Proceedings of the Ninth International Workshop \non Treebanks and Linguistic Theories. \nEditors: Markus Dickinson, Kaili Müürisep and Marco Passarotti. \nNEALT Proceedings Series, Vol. 9 (2010), 7-18. \n© 2010 The editors and contributors. \nPublished by \nNorthern European Association for Language \nTechnology (NEALT) \nhttp://omilia.uio.no/nealt. \nElectronically published at \nTartu University Library (Estonia) \nhttp://hdl.handle.net/10062/15891.
We describe a new test for unfamiliar face matching, the Glasgow Face Matching Test (GFMT). Viewers are shown pairs of faces, photographed in full-face view but with different cameras, and are asked to make same/different judgments. The full version of the test comprises 168 face pairs, and we also describe a shortened version with 40 pairs. We provide normative data for these tests derived from large subject samples. We also describe associations between the GFMT and other tests of matching and memory. The new test correlates moderately with face memory but more strongly with object matching, a result that is consistent with previous research highlighting a link between object and face matching, specific to unfamiliar faces. The test is available free for scientific use.
This paper demonstrates the implementation of a small monolingual lexical database for German (currently 7000 entries) for the purpose of manual and automated lexical queries. The lexicon is part of a web based text analysis application which serves to systematically analyze patient’s narratives from psychotherapy sessions. The narratives, small stories from everyday life, are conceived as dramaturgically constructed and performed linguistic productions (Boothe, 2004). The specific function of the lexicon in this context is to provide the means for the lexical coding of the story vocabulary, which is an important step in the narrative analysis1. The interdisciplinary project is located in psychology, but involves likewise linguistics, corpus linguistics, lexicography, interactional linguistics and conversation analysis.
On the question of precisely what role common sense (or related datum like folk psychology, trust in pre-theoretic/intuitive judgments, etc.) should have in reigning in the possible excesses of our philosophical methods, the so-called 'continental' answer to this question, for the vast majority, would be "as little as possible", whereas the analytic answer for the vast majority would be "a reasonably central one". While this difference at the level of both rhetoric and meta-philosophy is sometimes -perhaps often -problematised by the actual philosophical practices of representative philosophers of either tradition, I will argue that this norm (and its absence) nonetheless continues to play an important justificatory role in relation to the use of some rather different methodological practices. In particular, many analytic philosophers not only explicitly invoke the value of common sense, but they also implicitly value it via techniques like conceptual analysis that want to explicate folk psychology and/or lay bare what is already embedded in the linguistic norms of a given culture, the widespread use of thought experiments and the way they function as 'intuition pumps', as well as the general aim to achieve 'reflective equilibrium' between our intuitions and reflective judgments in epistemology and political philosophy. Such methods, I will argue, enshrine a conservative, or, more positively, a modest understanding of the philosophical project in that it is invested in cohering with both a given body of knowledge and common sense. These methods are notably less perspicuous in continental philosophy. To bring some of the reasons why this might be so to the fore, this paper considers Deleuze's sustained attack on both good and common sense, which he argues are fundamental to the prevalence of a dogmatic image of thought. If Deleuze is right about this, and if the analytic tradition distils and perfects certain methods that are closely associated with this image of thought, then we have here a rather stark methodological contrast that calls for elaboration and evaluation.