The problem is simple to state, and hard to do well: how does an interview transcript become a story? I keep coming back to that question because transcripts look fixed on the page, yet their meaning shifts as soon as we read them for plot, voice, and sequence.
A transcript is not a story by itself. It is raw evidence from speech. It gives us words, pauses, and sometimes fragments of exchange. What it does not give us is shape. Shape has to be made.
That is the central lesson I take from work on narrative analysis. The task is not to strip interviews down until only themes remain. The task is to build a readable account from speech that was never meant to arrive in finished form.
I find this useful because it names the real labor of qualitative writing. People often talk as if analysis begins after transcription. In practice, the first choices already matter at the moment speech is turned into text. What gets kept? What gets cleaned up? What gets marked as a pause, a tone, or a gesture? Those choices affect what can be seen later.
This is why transcription is part of interpretation, even before coding starts. If a transcript records only words, some meaning drops out. If it also captures tone, emphasis, or movement, the record changes. Neither version is neutral. Each one guides the reader toward a different understanding of the same exchange.
I think beginners often expect one clean method. They want a route from recording to result. But narrative work does not move that way. It moves in steps, and the steps loop back on each other. A researcher reads, re-reads, sorts, and revises. The story often appears only after several passes.
One practical way to think about this is to ask four questions in order. First, what happened in the interview? Second, what parts of the account belong together in time? Third, what details need support from a follow-up exchange? Fourth, what larger meaning emerges when the pieces are joined? Those questions do not solve the work for you. They give it a shape.
Here is a small example. Suppose a participant says, “I left school, worked nights, and only came back when my sister told me I could.” A bare thematic code might note “education interruption.” A story-based reading asks for more. Did the speaker leave because of money, care work, fear, or something else? Why does the sister matter? What changed on the way back? The point is not to decorate the quote. The point is to place it in time and in relation to other events.
That is also where the difference between analysis and interpretation helps. Analysis is the close work of sorting parts. Interpretation is the broader act of saying what those parts mean together. The two blur in practice. Still, the distinction matters because it reminds us that a transcript can be read at more than one level.
For me, the most honest part of this method is its dependence on judgment. No fixed recipe turns interviews into good narrative account. The data do not contain a hidden story waiting to be uncovered in one pass. Meaning is made through careful decisions. Those decisions should be plain enough that another reader can see the path.
I also value the collaborative side of this work. A follow-up conversation can clear up a missing point or confirm a sequence that seemed uncertain in the first transcript. That matters because a good story from interview data is not built by force. It is built by checking what was meant, what was said, and what was left unsaid.
This is where a database or digital collection lesson enters my own work. Search tools often make text feel stable. They return words, not context. That can be useful, but it can also hide the fact that a transcript is already a shaped object. If we want to search and study it well, we need to remember how much editorial work sits inside the file before analysis begins.
A useful search tip follows from that. When dealing with interview material, search for both content and form. Look for terms tied to events, but also for terms tied to method, like transcription, restorying, emplotment, or narrative analysis. Those words tell you how the source was made, and that often matters as much as what it says.
The honest limit is also clear. A transcript cannot fully recover voice, face, or setting. It can only point back to them. That means every story built from interview text is a careful reconstruction, not a perfect replay. Good practice does not erase that gap. It keeps the gap visible.
What I learn from this kind of work is plain enough. I can now explain how a transcript becomes a narrative account, where the key choices sit, and why those choices belong to analysis as much as to writing. I can also see why different readers may build different stories from the same interview set, and why that is a feature of the method, not a failure of it.
That is the kind of plain work I want The Source List to keep doing: one digital source worth knowing, one search tip, and one honest limitation, stated without varnish.