A Boat Is a Boat Is a Boat…Unless It Is a Horse – Rethinking the Role of Typology
Today, it is widely accepted that typology is a biased and inconsistent attempt to classify archaeological material based on the similarity of a predefined set of features. In this respect, machine learning (ML) works similar to typology. ML approaches are often deployed because it is thought that t...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
De Gruyter
2022-12-01
|
Series: | Open Archaeology |
Subjects: | |
Online Access: | https://doi.org/10.1515/opar-2022-0277 |
_version_ | 1797947790961147904 |
---|---|
author | Horn Christian Green Ashely Skärström Victor Wåhlstrand Lindhé Cecilia Peternell Mark Ling Johan |
author_facet | Horn Christian Green Ashely Skärström Victor Wåhlstrand Lindhé Cecilia Peternell Mark Ling Johan |
author_sort | Horn Christian |
collection | DOAJ |
description | Today, it is widely accepted that typology is a biased and inconsistent attempt to classify archaeological material based on the similarity of a predefined set of features. In this respect, machine learning (ML) works similar to typology. ML approaches are often deployed because it is thought that they reduce biases. However, biases are introduced into the process at many points, e.g., feature selection. In a project applying ML to Scandinavian rock art data, it was noticed that the algorithm struggles with classifying certain motifs correctly. This contribution discusses the consistency in applying biases by ML in contrast to the inconsistency of human classification. It is argued that it is necessary to bring machines and humans into a meaningful dialogue attempting to understand why apparent “misclassifications” happen. This is important to inform us about the classification output, our biases, and the rock art data, which are in themself inconsistent, ambiguous, and biased because they are the outcomes of human creativity. The human inconsistency is a necessary component because in rock art not everything that looks similar has a similar meaning. |
first_indexed | 2024-04-10T21:32:16Z |
format | Article |
id | doaj.art-ff490e65103c4bc08d7bafae013fe16e |
institution | Directory Open Access Journal |
issn | 2300-6560 |
language | English |
last_indexed | 2024-04-10T21:32:16Z |
publishDate | 2022-12-01 |
publisher | De Gruyter |
record_format | Article |
series | Open Archaeology |
spelling | doaj.art-ff490e65103c4bc08d7bafae013fe16e2023-01-19T13:20:28ZengDe GruyterOpen Archaeology2300-65602022-12-01811218123010.1515/opar-2022-0277A Boat Is a Boat Is a Boat…Unless It Is a Horse – Rethinking the Role of TypologyHorn Christian0Green Ashely1Skärström Victor Wåhlstrand2Lindhé Cecilia3Peternell Mark4Ling Johan5Department of Historical Studies, Swedish Rock Art Research Archives, University of Gothenburg, Box 200, 405 30, Gothenburg, SwedenDepartment of Historical Studies, Swedish Rock Art Research Archives, University of Gothenburg, Box 200, 405 30, Gothenburg, SwedenDepartment of Literature, Centre for Digital Humanities, History of Ideas, and Religion, University of Gothenburg, Box 200, 405 30, Gothenburg, SwedenDepartment of Literature, History of Ideas, and Religion, Centre for Digital Humanities, University of Gothenburg, Box 200, 405 30, Gothenburg, SwedenDepartment of Earth Sciences, University of Gothenburg, Box 200, 405 30, Gothenburg, SwedenDepartment of Historical Studies, Swedish Rock Art Research Archives, University of Gothenburg, Box 200, 405 30, Gothenburg, SwedenToday, it is widely accepted that typology is a biased and inconsistent attempt to classify archaeological material based on the similarity of a predefined set of features. In this respect, machine learning (ML) works similar to typology. ML approaches are often deployed because it is thought that they reduce biases. However, biases are introduced into the process at many points, e.g., feature selection. In a project applying ML to Scandinavian rock art data, it was noticed that the algorithm struggles with classifying certain motifs correctly. This contribution discusses the consistency in applying biases by ML in contrast to the inconsistency of human classification. It is argued that it is necessary to bring machines and humans into a meaningful dialogue attempting to understand why apparent “misclassifications” happen. This is important to inform us about the classification output, our biases, and the rock art data, which are in themself inconsistent, ambiguous, and biased because they are the outcomes of human creativity. The human inconsistency is a necessary component because in rock art not everything that looks similar has a similar meaning.https://doi.org/10.1515/opar-2022-0277machine learningsegmentationtypologysimilarityhuman creativityrock art |
spellingShingle | Horn Christian Green Ashely Skärström Victor Wåhlstrand Lindhé Cecilia Peternell Mark Ling Johan A Boat Is a Boat Is a Boat…Unless It Is a Horse – Rethinking the Role of Typology Open Archaeology machine learning segmentation typology similarity human creativity rock art |
title | A Boat Is a Boat Is a Boat…Unless It Is a Horse – Rethinking the Role of Typology |
title_full | A Boat Is a Boat Is a Boat…Unless It Is a Horse – Rethinking the Role of Typology |
title_fullStr | A Boat Is a Boat Is a Boat…Unless It Is a Horse – Rethinking the Role of Typology |
title_full_unstemmed | A Boat Is a Boat Is a Boat…Unless It Is a Horse – Rethinking the Role of Typology |
title_short | A Boat Is a Boat Is a Boat…Unless It Is a Horse – Rethinking the Role of Typology |
title_sort | boat is a boat is a boat unless it is a horse rethinking the role of typology |
topic | machine learning segmentation typology similarity human creativity rock art |
url | https://doi.org/10.1515/opar-2022-0277 |
work_keys_str_mv | AT hornchristian aboatisaboatisaboatunlessitisahorserethinkingtheroleoftypology AT greenashely aboatisaboatisaboatunlessitisahorserethinkingtheroleoftypology AT skarstromvictorwahlstrand aboatisaboatisaboatunlessitisahorserethinkingtheroleoftypology AT lindhececilia aboatisaboatisaboatunlessitisahorserethinkingtheroleoftypology AT peternellmark aboatisaboatisaboatunlessitisahorserethinkingtheroleoftypology AT lingjohan aboatisaboatisaboatunlessitisahorserethinkingtheroleoftypology AT hornchristian boatisaboatisaboatunlessitisahorserethinkingtheroleoftypology AT greenashely boatisaboatisaboatunlessitisahorserethinkingtheroleoftypology AT skarstromvictorwahlstrand boatisaboatisaboatunlessitisahorserethinkingtheroleoftypology AT lindhececilia boatisaboatisaboatunlessitisahorserethinkingtheroleoftypology AT peternellmark boatisaboatisaboatunlessitisahorserethinkingtheroleoftypology AT lingjohan boatisaboatisaboatunlessitisahorserethinkingtheroleoftypology |