Theoretical Virtues in Eighteenth-Century Debates on Animal Cognition.
Van den Berg, H.
August 2019.
link
bibtex
abstract
@misc{van_den_berg_theoretical_2019,
address = {Prague},
type = {Contributed talk},
title = {Theoretical {Virtues} in {Eighteenth}-{Century} {Debates} on {Animal} {Cognition}},
abstract = {Within eighteenth-century debates on animal cognition we can distinguish at least three main theoretical positions: (i) Buffon’s mechanism, (ii) Reimarus’ theory of instincts, and (iii) the sensationalism of Condillac and Leroy. In this paper, I will provide a precise analysis and reconstruction of the empirical and theoretical reasons these authors gave for adopting these positions. I will argue that in order to fully understand why Buffon, Reimarus, Condillac, and Leroy adopted their positions, we must pay special attention to the theoretical virtues these naturalists alluded to while justifying their position. These theoretical virtues have received little to no attention in the literature on eighteenth-century animal cognition, but figure prominently in the justification of the mechanist, instinctive, and sensationalist theories of animal behavior. In addition, I will provide a precise analysis of the debate on the explanatory scope of mechanism in eighteenth-century animal cognition, which is central to understanding the three positions I have outlined.},
author = {Van den Berg, Hein},
month = aug,
year = {2019},
}
Within eighteenth-century debates on animal cognition we can distinguish at least three main theoretical positions: (i) Buffon’s mechanism, (ii) Reimarus’ theory of instincts, and (iii) the sensationalism of Condillac and Leroy. In this paper, I will provide a precise analysis and reconstruction of the empirical and theoretical reasons these authors gave for adopting these positions. I will argue that in order to fully understand why Buffon, Reimarus, Condillac, and Leroy adopted their positions, we must pay special attention to the theoretical virtues these naturalists alluded to while justifying their position. These theoretical virtues have received little to no attention in the literature on eighteenth-century animal cognition, but figure prominently in the justification of the mechanist, instinctive, and sensationalist theories of animal behavior. In addition, I will provide a precise analysis of the debate on the explanatory scope of mechanism in eighteenth-century animal cognition, which is central to understanding the three positions I have outlined.
Computational methods for the history of philosophy: Interpretative models and corpus-building.
Ginammi, A.; and Sommerauer, P.
November 2019.
Paper
link
bibtex
abstract
1 download
@misc{ginammi_computational_2019,
title = {Computational methods for the history of philosophy: {Interpretative} models and corpus-building},
url = {https://docs.google.com/presentation/d/1sWTwE0xoF5jOdkS7uG5EuaWxD2xKyXfWd7mYJ2I9-No/edit#slide=id.g6b2111c0b8_0_0},
abstract = {In what way can computational and quantitative methods supplement the largely qualitative methods of
traditional history of philosophy? In this talk, we will give some answers to that question, by means of
presenting several projects in the history of philosophy using computational methods carried out within
our team. The methods applied in these projects are rather different and range from count-based, highly
basic computational techniques, to rather sophisticated information retrieval tools based on
distributional semantics, to the use of formal ontologies. What these projects have in common is that
they use novel methods and thereby urge us to reflect on the methodology: how can these methods be
used in a transparent and reliable manner?
The majority of our projects employs the ``model-approach” which we introduced in (removed for blind
review), i.e. they use explicit conceptual frameworks which facilitate the interpretation of texts. We
argue that the model approach is a useful method to obtain philosophically relevant information from
the data delivered by quantitative computational methods, and at the same time makes the
interpretation of this data by researchers less arbitrary and biased. One of our projects supplements this
model approach by defined annotation schemes to capture differences and similarities between various
conceptions in a way which is as accountable as possible.
An undeniable benefit of computational methods is that research can be done on a scale which is
unthinkable for traditional methods. However, this raises a new issue which, in our view, hitherto
received insufficient attention: how to build a representative corpus relative to a given historical
research question? We discuss our efforts to build our datasets as systematically as possible, and
propose that corpus-building for historical-interpretive research should rely on a method taking
inspiration from the so-called “systematic literature review” common in the sciences. Additionally, we
will discuss the issues we encountered by building multilingual, multi-script corpora for our
computational research.},
author = {Ginammi, Annapaola and Sommerauer, Pia},
collaborator = {Oortwijn, Yvette and Meyer, François and Betti, Arianna and Fokkens, Antske and Bloem, Jelke},
month = nov,
year = {2019},
}
In what way can computational and quantitative methods supplement the largely qualitative methods of traditional history of philosophy? In this talk, we will give some answers to that question, by means of presenting several projects in the history of philosophy using computational methods carried out within our team. The methods applied in these projects are rather different and range from count-based, highly basic computational techniques, to rather sophisticated information retrieval tools based on distributional semantics, to the use of formal ontologies. What these projects have in common is that they use novel methods and thereby urge us to reflect on the methodology: how can these methods be used in a transparent and reliable manner? The majority of our projects employs the ``model-approach” which we introduced in (removed for blind review), i.e. they use explicit conceptual frameworks which facilitate the interpretation of texts. We argue that the model approach is a useful method to obtain philosophically relevant information from the data delivered by quantitative computational methods, and at the same time makes the interpretation of this data by researchers less arbitrary and biased. One of our projects supplements this model approach by defined annotation schemes to capture differences and similarities between various conceptions in a way which is as accountable as possible. An undeniable benefit of computational methods is that research can be done on a scale which is unthinkable for traditional methods. However, this raises a new issue which, in our view, hitherto received insufficient attention: how to build a representative corpus relative to a given historical research question? We discuss our efforts to build our datasets as systematically as possible, and propose that corpus-building for historical-interpretive research should rely on a method taking inspiration from the so-called “systematic literature review” common in the sciences. Additionally, we will discuss the issues we encountered by building multilingual, multi-script corpora for our computational research.
On how perspectival hylemorphism got the facts wrong.
Parisi, M. C.
Master's thesis, University of Amsterdam, Amsterdam, 2019.
DIAMOND \textbar Supervisor: Arianna Betti; Second Reader: Hein van den Berg
link
bibtex
@mastersthesis{parisi_how_2019,
address = {Amsterdam},
title = {On how perspectival hylemorphism got the facts wrong},
school = {University of Amsterdam},
author = {Parisi, Maria Chiara},
year = {2019},
note = {DIAMOND {\textbar} Supervisor: Arianna Betti; Second Reader: Hein van den Berg},
}
Quine's Naturalistic Epistemology; A Quantitative Investigation.
Ossenkoppele, T.
Ph.D. Thesis, University of Amsterdam, Amsterdam, 2019.
DIAMOND \textbar Supervisor: Arianna Betti
link
bibtex
@phdthesis{ossenkoppele_quines_2019,
address = {Amsterdam},
type = {{BA} thesis},
title = {Quine's {Naturalistic} {Epistemology}; {A} {Quantitative} {Investigation}},
school = {University of Amsterdam},
author = {Ossenkoppele, Thijs},
year = {2019},
note = {DIAMOND {\textbar} Supervisor: Arianna Betti},
}
Quine's Naturalistic Epistemology; a Quantitative Investigation.
Ossenkoppele, T.
November 2019.
Paper
link
bibtex
abstract
@misc{ossenkoppele_quines_2019-1,
type = {Invited {Lecture}},
title = {Quine's {Naturalistic} {Epistemology}; a {Quantitative} {Investigation}},
url = {https://docs.google.com/presentation/d/15bEYIRK-0PKjCnOqd9deDqQvGs7x3tk2SqIKsCAiUMg/edit?usp=sharing},
abstract = {In this presentation, I offer a supplementary, quantitative element to the close-reading based methods with which large-scale historico-philosophical research, aimed at confirming or disconfirming hypotheses about a certain historical philosopher's adherence to position x or y, is traditionally conducted. The conclusions which result from such research are often drawn on the basis of incomplete or selective evidence, and without the researcher's conceptual presuppositions having been made explicit. In order to be justified in ascribing either position x or position y to a historical philosopher, or in pointing out an inconsistent moving back and forth between x and y in that philosopher's work, the researcher ought to have taken all relevant passages concerning these positions into consideration. I argue that by supplementing traditional research with the use of interpretive models, in which the researcher's conceptual presuppositions have been made explicit, to guide the interpreting of passages found with computational tools in complete, high-quality corpora, historico-philosophical research can yield results based on a more comprehensive amount of evidence when compared with traditional methodology. At the same time, the use of complete corpora prevents the researcher from conveniently picking passages that support their hypotheses; all relevant data will have to be taken into consideration. In support of the functionality of this novel methodology, I present my application of it in answering a question which was raised by Susan Haack, and which was treated by both her and Sander Verhaegh in a traditional manner. The question concerns W.V.O. Quine's naturalistic epistemology: did he believe that a naturalized epistemology ought to be carried out from within the natural sciences alone (scientistic naturalistic epistemology), from within science in a broader sense (modest naturalistic epistemology), or was he ambivalent about the matter? Haack accused Quine of moving between modest and scientistic naturalism in his writings about epistemology, which would be problematic since the latter believed some of the epistemological questions that cannot be answered from within a scientistic naturalistic outlook are legitimate questions, such as: “do the natural sciences enjoy a privileged epistemic status?”. The evidence that Haack provided in support of her hypothesis is inconclusive, and the background assumptions she makes regarding notions such as 'natural science' were not made explicit. It is shown that with the application of the supplementary methods, thus using a model to help interpret passages concerning the methodology proper to epistemology from a corpus consisting of all of Quine's publications, Haack's question concerning the nature of Quine's naturalistic epistemology is answered with more conclusiveness and rigor than could be provided by traditional close reading-based research alone.},
author = {Ossenkoppele, Thijs},
month = nov,
year = {2019},
}
In this presentation, I offer a supplementary, quantitative element to the close-reading based methods with which large-scale historico-philosophical research, aimed at confirming or disconfirming hypotheses about a certain historical philosopher's adherence to position x or y, is traditionally conducted. The conclusions which result from such research are often drawn on the basis of incomplete or selective evidence, and without the researcher's conceptual presuppositions having been made explicit. In order to be justified in ascribing either position x or position y to a historical philosopher, or in pointing out an inconsistent moving back and forth between x and y in that philosopher's work, the researcher ought to have taken all relevant passages concerning these positions into consideration. I argue that by supplementing traditional research with the use of interpretive models, in which the researcher's conceptual presuppositions have been made explicit, to guide the interpreting of passages found with computational tools in complete, high-quality corpora, historico-philosophical research can yield results based on a more comprehensive amount of evidence when compared with traditional methodology. At the same time, the use of complete corpora prevents the researcher from conveniently picking passages that support their hypotheses; all relevant data will have to be taken into consideration. In support of the functionality of this novel methodology, I present my application of it in answering a question which was raised by Susan Haack, and which was treated by both her and Sander Verhaegh in a traditional manner. The question concerns W.V.O. Quine's naturalistic epistemology: did he believe that a naturalized epistemology ought to be carried out from within the natural sciences alone (scientistic naturalistic epistemology), from within science in a broader sense (modest naturalistic epistemology), or was he ambivalent about the matter? Haack accused Quine of moving between modest and scientistic naturalism in his writings about epistemology, which would be problematic since the latter believed some of the epistemological questions that cannot be answered from within a scientistic naturalistic outlook are legitimate questions, such as: “do the natural sciences enjoy a privileged epistemic status?”. The evidence that Haack provided in support of her hypothesis is inconclusive, and the background assumptions she makes regarding notions such as 'natural science' were not made explicit. It is shown that with the application of the supplementary methods, thus using a model to help interpret passages concerning the methodology proper to epistemology from a corpus consisting of all of Quine's publications, Haack's question concerning the nature of Quine's naturalistic epistemology is answered with more conclusiveness and rigor than could be provided by traditional close reading-based research alone.
The semantics of meaning: distributional approaches for studying philosophical text [ongoing].
Oortwijn, Y.; and Meyer, F.
Technical Report 2019.
link
bibtex
@techreport{oortwijn_semantics_2019,
title = {The semantics of meaning: distributional approaches for studying philosophical text [ongoing]},
author = {Oortwijn, Yvette and Meyer, François},
year = {2019},
}
Interacting with open collections: Ariadne, BolVis and KantVis.
Ginammi, A.; and Betti, A.
June 2019.
link
bibtex
@misc{ginammi_interacting_2019,
type = {Invited {Talk}},
title = {Interacting with open collections: {Ariadne}, {BolVis} and {KantVis}},
author = {Ginammi, Annapaola and Betti, Arianna},
month = jun,
year = {2019},
}
The e-Ideas project: Modeling meaning and change in the history of ideas.
Bloem, J.
September 2019.
tex.ids: bloemEIdeasProjectModeling2019a
link
bibtex
@misc{bloem_e-ideas_2019,
address = {Berlin},
title = {The e-{Ideas} project: {Modeling} meaning and change in the history of ideas},
author = {Bloem, Jelke},
month = sep,
year = {2019},
note = {tex.ids: bloemEIdeasProjectModeling2019a},
}
Hoe leert een computer een taal?.
Bloem, J.
October 2019.
link
bibtex
@misc{bloem_hoe_2019,
address = {Amsterdam},
title = {Hoe leert een computer een taal?},
author = {Bloem, Jelke},
month = oct,
year = {2019},
}
Building Corpora: Objectivity and Quality of Corpora for Philosophical Research.
Betti, A.; and Ginammi, A.
February 2019.
tex.ids: ginammiBuildingCorporaObjectivity2019
link
bibtex
@misc{betti_building_2019,
address = {Torino, Italy},
type = {Invited {Talk}},
title = {Building {Corpora}: {Objectivity} and {Quality} of {Corpora} for {Philosophical} {Research}},
author = {Betti, Arianna and Ginammi, Annapaola},
month = feb,
year = {2019},
note = {tex.ids: ginammiBuildingCorporaObjectivity2019},
}
Bolzano's theory of ground and consequence and the traditional theory of concepts.
Ginammi, A.; and Betti, A.
August 2019.
Paper
link
bibtex
@misc{ginammi_bolzanos_2019,
address = {Prague},
type = {Contributed talk},
title = {Bolzano's theory of ground and consequence and the traditional theory of concepts},
url = {https://www.overleaf.com/read/sqxgcgssqxmr},
author = {Ginammi, Annapaola and Betti, Arianna},
month = aug,
year = {2019},
}
Evaluating the consistency of word embeddings from small data.
Bloem, J.; Fokkens, A.; and Herbelot, A.
January 2019.
tex.ids: bloemEvaluatingConsistencyWord2019b
link
bibtex
@misc{bloem_evaluating_2019,
address = {Utrecht},
title = {Evaluating the consistency of word embeddings from small data},
author = {Bloem, Jelke and Fokkens, Antske and Herbelot, Aurélie},
month = jan,
year = {2019},
note = {tex.ids: bloemEvaluatingConsistencyWord2019b},
}
Learning consistent embeddings from small data with Nonce2Vec.
Bloem, J.
March 2019.
link
bibtex
@misc{bloem_learning_2019,
address = {Amsterdam},
title = {Learning consistent embeddings from small data with {Nonce2Vec}},
author = {Bloem, Jelke},
month = mar,
year = {2019},
}
e-Ideas: Naar een computationele ideeëngeschiedenis.
Bloem, J.
March 2019.
link
bibtex
@misc{bloem_e-ideas_2019-1,
address = {Amsterdam},
title = {e-{Ideas}: {Naar} een computationele ideeëngeschiedenis},
author = {Bloem, Jelke},
month = mar,
year = {2019},
}
Knowledge, Trust, Technology.
Betti, A.
November 2019.
link
bibtex
@misc{betti_knowledge_2019,
address = {Università di Bologna},
type = {Invited {Talk}},
title = {Knowledge, {Trust}, {Technology}},
author = {Betti, Arianna},
month = nov,
year = {2019},
}
(Non)Symbolic AI in Computational History of Philosophy.
Betti, A.
November 2019.
link
bibtex
@misc{betti_nonsymbolic_2019,
address = {VU University, Amsterdam},
type = {Invited {Talk}},
title = {({Non}){Symbolic} {AI} in {Computational} {History} of {Philosophy}},
author = {Betti, Arianna},
month = nov,
year = {2019},
}
‘Trust in VIS4DH’ Expert Panel Pitch.
Betti, A.
October 2019.
link
bibtex
@misc{betti_trust_2019,
address = {Vancouver, Canada (CA)},
type = {Invited {Talk}},
title = {‘{Trust} in {VIS4DH}’ {Expert} {Panel} {Pitch}},
author = {Betti, Arianna},
month = oct,
year = {2019},
}
Domain Extension and Ideal Elements in Mathematics.
Bellomo, A.
May 2019.
Paper
link
bibtex
1 download
@misc{bellomo_domain_2019,
address = {Salzburg},
type = {Invited talk},
title = {Domain {Extension} and {Ideal} {Elements} in {Mathematics}},
url = {https://www.uni-salzburg.at/index.php?id=66639&L=1},
author = {Bellomo, Anna},
month = may,
year = {2019},
}
Dedekind and Manders: two conceptions of domain extension?.
Bellomo, A.
January 2019.
link
bibtex
@misc{bellomo_dedekind_2019,
address = {Cambridge, UK},
type = {Contributed talk},
title = {Dedekind and {Manders}: two conceptions of domain extension?},
author = {Bellomo, Anna},
month = jan,
year = {2019},
}
Evaluating the Consistency of Word Embeddings from Small Data.
Bloem, J.; Fokkens, A.; and Herbelot, A.
In
Proceedings of the International Conference Recent Advances in Natural Language Processing, RANLP 2019, pages 132–141, Varna, Bulgaria, 2019. Incoma Ltd
Paper
link
bibtex
@inproceedings{bloem_evaluating_2019-1,
address = {Varna, Bulgaria},
title = {Evaluating the {Consistency} of {Word} {Embeddings} from {Small} {Data}},
url = {http://lml.bas.bg/ranlp2019/proceedings-ranlp-2019.pdf},
booktitle = {Proceedings of the {International} {Conference} {Recent} {Advances} in {Natural} {Language} {Processing}, {RANLP} 2019},
publisher = {Incoma Ltd},
author = {Bloem, Jelke and Fokkens, Antske and Herbelot, Aurelie},
year = {2019},
pages = {132--141},
}
History of Philosophy in Ones and Zeros.
Betti, A.; van den Berg, H.; Oortwijn, Y.; and Treijtel, C.
In Curtis, M.; and Fischer, E., editor(s),
Methodological Advances in Experimental Philosophy, of Advances in Experimental Philosophy, pages 295–332. Bloomsbury, London, March 2019.
GREEN
Paper
link
bibtex
abstract
16 downloads
@incollection{betti_history_2019,
address = {London},
series = {Advances in {Experimental} {Philosophy}},
title = {History of {Philosophy} in {Ones} and {Zeros}},
url = {https://1drv.ms/b/s!AgPq3zEkkYuOiP5Q8-ia4_sSI3CyOw},
abstract = {How can we best reconstruct the origin of a notion, its development, and possible spread to multiple fields? We present a pilot study on the spread of the notion of conceptual scheme. Though the notion is philosophically important, its origin, development, and spread are unclear. Several purely qualitative and competing historical hypotheses have been offered, which rely on disconnected disciplinary traditions, and have never been tested all at once in a single comprehensive investigation fitting the scope of its subject matter. As a step toward such an investigation, we trace the use of the bigram “conceptual scheme” in about 42,000 US journal articles in social sciences from 1888-1959 by using a novel method combining a quantitative procedure aided by basic computational techniques with qualitative elements informed by Betti and van den Berg (2014)’s ‘model approach to the history of ideas’.},
booktitle = {Methodological {Advances} in {Experimental} {Philosophy}},
publisher = {Bloomsbury},
author = {Betti, Arianna and van den Berg, Hein and Oortwijn, Yvette and Treijtel, Caspar},
editor = {Curtis, Mark and Fischer, Eugen},
month = mar,
year = {2019},
note = {GREEN},
pages = {295--332},
}
How can we best reconstruct the origin of a notion, its development, and possible spread to multiple fields? We present a pilot study on the spread of the notion of conceptual scheme. Though the notion is philosophically important, its origin, development, and spread are unclear. Several purely qualitative and competing historical hypotheses have been offered, which rely on disconnected disciplinary traditions, and have never been tested all at once in a single comprehensive investigation fitting the scope of its subject matter. As a step toward such an investigation, we trace the use of the bigram “conceptual scheme” in about 42,000 US journal articles in social sciences from 1888-1959 by using a novel method combining a quantitative procedure aided by basic computational techniques with qualitative elements informed by Betti and van den Berg (2014)’s ‘model approach to the history of ideas’.
SolarView: Low Distortion Radial Embedding with a Focus.
Castermans, T.; Verbeek, K.; Speckmann, B.; Westenberg, M. A.; Koopman, R.; Wang, S.; van den Berg, H.; and Betti, A.
IEEE transactions on visualization and computer graphics, 25(10): 2969–2982. October 2019.
DIAMOND
doi
link
bibtex
abstract
@article{castermans_solarview_2019,
title = {{SolarView}: {Low} {Distortion} {Radial} {Embedding} with a {Focus}},
volume = {25},
issn = {1077-2626},
doi = {10.1109/TVCG.2018.2865361},
abstract = {We propose a novel type of low distortion radial embedding which focuses on one specific entity and its closest neighbors. Our embedding preserves near-exact distances to the focus entity and aims to minimize distortion between the other entities. We present an interactive exploration tool SolarView which places the focus entity at the center of a “solar system” and embeds its neighbors guided by concentric circles. SolarView provides an implementation of our novel embedding and several state-of-the-art dimensionality reduction and embedding techniques, which we adapted to our setting in various ways. We experimentally evaluated our embedding and compared it to these state-of-the-art techniques. The results show that our embedding competes with these techniques and achieves low distortion in practice. Our method performs particularly well when the visualization, and hence the embedding, adheres to the solar system design principle of our application. Nonetheless-as with all dimensionality reduction techniques-the distortion may be high. We leverage interaction techniques to give clear visual cues that allow users to accurately judge distortion. We illustrate the use of SolarView by exploring the high-dimensional metric space of bibliographic entity similarities.},
language = {English},
number = {10},
journal = {IEEE transactions on visualization and computer graphics},
author = {Castermans, Thom and Verbeek, Kevin and Speckmann, Bettina and Westenberg, Michel A. and Koopman, Rob and Wang, Shenghui and van den Berg, Hein and Betti, Arianna},
month = oct,
year = {2019},
note = {DIAMOND},
keywords = {Dimensionality reduction, Radial embedding, Visualizing distortion},
pages = {2969--2982},
}
We propose a novel type of low distortion radial embedding which focuses on one specific entity and its closest neighbors. Our embedding preserves near-exact distances to the focus entity and aims to minimize distortion between the other entities. We present an interactive exploration tool SolarView which places the focus entity at the center of a “solar system” and embeds its neighbors guided by concentric circles. SolarView provides an implementation of our novel embedding and several state-of-the-art dimensionality reduction and embedding techniques, which we adapted to our setting in various ways. We experimentally evaluated our embedding and compared it to these state-of-the-art techniques. The results show that our embedding competes with these techniques and achieves low distortion in practice. Our method performs particularly well when the visualization, and hence the embedding, adheres to the solar system design principle of our application. Nonetheless-as with all dimensionality reduction techniques-the distortion may be high. We leverage interaction techniques to give clear visual cues that allow users to accurately judge distortion. We illustrate the use of SolarView by exploring the high-dimensional metric space of bibliographic entity similarities.