The short version, from a reference in the article.
“The primary purpose of narrative,” media scholar Katherine Hayles argued several years ago, “is to search for meaning,” which makes “narrative an essential technology for human beings, who can arguably be defined as meaning-seeking animals.”
His "narrative vs database" essay talks around the problem. Narratives tend to discuss causation, either explicitly or by implication. Raw data does not contain causation information. That's the real distinction here.
We know that humans are hard-wired to find causation, even when it may not exist.[1] This has survival value. In a hostile environment, an excessive false alarm rate is better for survival than missing a threat.
If you do correlation on enough data, you find what looks like causation. Often it's just noise.
This is a well known phenomenon. Intelligence analysts, investment quants, and people who analyze research data have to be trained to watch for it. Most people don't have that kind of training.
Under information overload, this gets worse. Combine this with the human tendency to find causation when it doesn't exist, and you get false narratives. Even without wishful thinking or bias.
You appear to be strawmaning this piece to a large degree.
The necessity for humans to find meaning isn't ignored or argued as anything other than "how human brains work", in the piece.
The Database metaphor isn't a separation of raw data from narrative -- it's recognizing that in the modern zeitgeist, the abundance of data is so vast that a new experience has emerged that supersedes any Narrative, that's the Database: a super collection of all raw data as well as the known paths through it.
The existence of the Database then calls into question the validity of any one Narrative, and the rest of the piece follows.
“The primary purpose of narrative,” media scholar Katherine Hayles argued several years ago, “is to search for meaning,” which makes “narrative an essential technology for human beings, who can arguably be defined as meaning-seeking animals.”
His "narrative vs database" essay talks around the problem. Narratives tend to discuss causation, either explicitly or by implication. Raw data does not contain causation information. That's the real distinction here.
We know that humans are hard-wired to find causation, even when it may not exist.[1] This has survival value. In a hostile environment, an excessive false alarm rate is better for survival than missing a threat.
If you do correlation on enough data, you find what looks like causation. Often it's just noise. This is a well known phenomenon. Intelligence analysts, investment quants, and people who analyze research data have to be trained to watch for it. Most people don't have that kind of training.
Under information overload, this gets worse. Combine this with the human tendency to find causation when it doesn't exist, and you get false narratives. Even without wishful thinking or bias.
It's not mysterious. It's how human brains work.
[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4008651/pdf/nih...