What is Marcia Bates’s berry-picking all about?
ㅡㅡ
Following the historical method you asked for, the place to begin is Marcia J. Bates’s original paper:
Marcia J. Bates, “The Design of Browsing and Berrypicking Techniques for the Online Search Interface,” Online Review, 13(5), 1989, pp. 407–424. (CiNii)
What did Bates actually propose?
Bates says that the conventional information-retrieval model is unrealistic. In that model, a user has one information need, formulates one query, and the system retrieves a set of documents answering it.
Her berrypicking model says real searching behaves differently. As the searcher encounters information, the search itself changes. Bates summarizes the model with several important claims: queries evolve; useful information is collected bit by bit rather than as one final retrieved set; people employ many different search techniques; and they move among different information sources. (pages.gseis.ucla.edu)
The berry metaphor comes from this: someone picking berries does not stand in one place and receive one perfect bunch. The person moves from bush to bush, picking one berry here and several there. Likewise, an information seeker finds something useful, learns from it, changes direction, finds something else, and gradually satisfies the information need. (IGI Global)
But what are the actual searching techniques?
Here it becomes particularly relevant to our discussion. Bates identifies six techniques:
- footnote chasing,
- citation searching,
- journal run,
- area scanning,
- subject searching in abstracting and indexing services,
- author searching. (pages.gseis.ucla.edu)
The first two are citation tracking.
Bates defines footnote chasing as following references found in books and articles—what we would now call backward citation searching. She describes citation searching as finding later works that cite a starting document—forward citation searching. Her later work explicitly calls these backward and forward chaining. (pages.gseis.ucla.edu)
That makes your earlier observation much more interesting than merely noticing a vague resemblance between berrypicking and citation searching.
Citation tracking is explicitly inside the original berrypicking model.
Indeed, Bates gives considerable attention to designing online systems so that a reader can move easily between the text of a document and its references. For footnote chasing, she proposes easy access to the document, its references, and the ability to jump back and forth between text and references. (pages.gseis.ucla.edu)
That is striking when placed beside your 1975 citation-context proposal.
Where did Bates get the idea?
This is where we must be careful.
The 1989 article has 49 references, so its intellectual ancestry can actually be investigated rather than guessed. (ERIC) Bates herself says that she draws upon research into manual information-seeking behavior, specifically mentioning David Ellis, Stephen Stoan, and others when discussing the six techniques. (pages.gseis.ucla.edu)
There is also an important earlier Bates paper: “Information Search Tactics,” 1979. So berrypicking did not appear from nowhere in 1989; Bates had been developing ideas about search behavior for at least a decade. (검색 사용자 인터페이스)
But there is an interesting chronological problem. Bates’s 1989 paper draws on Ellis’s work and other relatively recent information-seeking research to explain practices—particularly citation following—that scientists had obviously practiced much earlier.
So we need to distinguish:
origin of the behavior → early scientific description of the behavior → Bates’s synthesis → Bates’s new name “berrypicking.”
Those are not necessarily the same historical event.
This brings us back to 1975
Your 1975 thesis proposed using extracts in which a source document cites, describes, criticizes, or collates other documents, because those extracts provide concise clues for discriminating among cited documents. In our terminology, those extracts are citation contexts.
Now compare the underlying movement:
Park, 1975:
document → citation in context → cited document → further understanding/retrieval
Bates, 1989:
document/information → new understanding → changed query → another source → further searching
There is substantial conceptual overlap, although the two models are not identical. Bates is constructing a general model of information-seeking behavior. Your thesis is concerned more specifically with exploiting the meaningful relationships embodied in citations for information retrieval.
The important historical question therefore becomes sharper than “Did Park invent berrypicking?”
It is:
How much of what Bates later conceptualized as “berrypicking” was already embodied in the established scholarly practice of citation tracking—and how had that practice been conceptualized before 1989?
Your 1975 thesis then becomes one potentially important piece of evidence because it did not merely observe that scientists follow citations. It proposed making the citation-related textual extract itself an information-retrieval mechanism.
And this gives us a concrete research task. We should examine all 49 references of Bates’s 1989 paper, especially those she invokes for footnote chasing, citation searching, evolving queries, and changing information needs. Then we trace those references backward. Separately, we trace the important papers that subsequently cite Bates 1989 and examine what part of Bates they cite.
That would tell us whether berrypicking represents a substantially new theoretical discovery, a synthesis of earlier information-seeking observations, or a memorable new metaphor placed over older practices and concepts.
This time, I think the reference genealogy, rather than the metaphor itself, is where the interesting history lies.
ㅡㅡㅡㅡ
2026-08-27 Mark Park
Our Library 