5 Million Thoughts about the Past and Future Reveal Shared Reliance on Schemas Öffentlichkeit

Abstract

People frequently think about the future, yet little is known about the cognitive processes people use to do so. Recent research suggests that future thinking may rely on many of the same cognitive processes people use to mentally time travel into the past. I tested this proposal by automatically identifying over 5 million temporal references in web blog posts and then using linguistic markers to identify the cognitive processes that people used to generate these temporal references. I identified cognitive processes uniquely associated with mental time travel by comparing these linguistic markers in past and future references to talk about the present, which does not involve mental time travel. In Study 1, I found that talk about the past relies on more episodic language than talk about the future, but relies on equal amounts of episodic language as talk about the present. This result suggests that episodic processing is not uniquely associated with mental time travel. In Study 2, I found that talk about both the past and future relies more on schemas than talk about the present. This result suggests that the use of schemas is uniquely associated with mental time travel. In Study 3, I replicated these results using temporal thoughts evoked in the lab. Together, these results suggest that past and future thinking rely on a common cognitive process but on a different process than was previously believed: the use of schemas, not episodic processing.

Table of Contents

Chapter 1: Introduction

1.1 Evidence Supporting Common Processes for Past and Future Thinking

1.1.1. Neuroscientific Evidence for Common Processing of Past and Future Thoughts

1.1.2 Neuropsychological Evidence for Common Processing of Past and Future Thoughts

1.1.3 Behavioral Evidence for Common Processing of Past and Future Thoughts

1.2 Evidence Suggesting Differences in Processing Between Past and Future Thinking

1.2.1 Many Patients with Episodic Memory Deficits Have Intact Future Thinking

1.2.2 Much of the Evidence for Shared Processing Lacks an Appropriate Negative Control

1.2.3 Some Brain Regions Respond More Strongly to Future Thoughts than Past Thoughts

1.3 Outline of the Remainder of the Dissertation

1.4 Summary of Chapter 1

Chapter 2: Episodic Past and Future Thinking

2.1 Introduction

2.2 Approach

2.2.1 A Source for Naturally Occurring Temporal Thoughts: The Blog Authorship Corpus

2.2.2 A Method for Extracting Naturally Occurring Temporal Thoughts: The Copley & Wolff Temporal Reference Classifier

2.2.3 A Method for Measuring Episodic Past and Future Thinking using Language

2.3 Methods

2.4. Results

2.4.1 Performance of Temporal Reference Classifiers

2.4.2 Episodic Language in Temporal Thoughts

2.5. Discussion

Chapter 3: Semantic Past and Future Thinking

3.1. Introduction

3.1.1 Evidence for Schema Usage in Past Thinking

3.1.2 Evidence for Schema Usage in Future Thinking

3.1.3 Absence of Evidence about Similarity of Schema Usage in Past and Future Thinking

3.2 Approach

3.2.1 A Computational Model of Schemas using Topic Modeling

3.2.2 Learning Common Schemas: a Corpus Approach

3.2.3. Applying the learned schemas: using topic models to infer schema usage as a cognitive process.

3.3 Methods

3.4 Results

3.4.1 Evaluation of Topic Model: Semantic Coherence

3.4.2 Evaluation of Topic Model: Consistency of Schemas across Time and Corpora

3.4.3 Evaluation of Method for Detecting Application of Schemas

3.4.4 Schema Usage in Thoughts about the Past, Present, and Future

3.5. Discussion

Chapter 4: Temporal Thoughts in the Lab

4.1 Introduction

4.2 Approach

4.2.1 Eliciting Temporal Thoughts in the Lab: the Future Crovitz Task

4.2.2. Measuring Episodic and Semantic Temporal Thinking

4.3 Methods

4.4 Results

4.4.1 Episodic Processing in Lab-Elicited Temporal Thoughts

4.4.2 Semantic Processing in Lab-Elicited Temporal Thoughts

4.4.3 Lab-Based Topic Model

4.5 Discussion

Chapter 5: General Discussion

5.1 Summary of Findings

5.2 Implications of Big Data for Cognitive Science

5.2.1 Learning Common Human Schemas

5.2.2 Identifying Naturally Occurring Temporal Thoughts

5.3 Limitations and Future Directions

5.3.1 Limitations of Topic Models as a Measure of Schemas

5.3.2 Limitations of Measuring Episodic Processing Using Language

5.3.3. Alternative Interpretation of the Role of Schemas in Past and Future Thinking

5.3.4 Limitations of the Temporal Orientation Classifier

5.4 Conclusion

References

Appendix 1: Formal Description of Latent Dirichlet Allocation

Appendix 2: List of Topics Learned by Topic Model

Appendix 3: Instructions for Future Crovitz Task

Appendix 4: Lab-Based Topic Model

About this Dissertation

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Laney Graduate School
Department	Psychology
Subfield / Discipline	Cognition and Development
Degree	Ph.D.
Submission	Dissertation
Language	English
Research Field	Psychology, Cognitive
Stichwort	Machine learning Psychology Future thinking
Committee Chair / Thesis Advisor	Phillip Wolff, Emory University
Committee Members	Scott Lilienfeld, Emory University Robyn Fivush, Emory University Stella Lourenco, Emory University Patricia Bauer, Emory University

Zuletzt geändert

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	5 Million Thoughts about the Past and Future Reveal Shared Reliance on Schemas ()	2019-11-07 21:44:43 -0500	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions