These haiku are derived from our papers, "Forgetting practices in the data sciences" (Muller & Strohmayer, 2022), "Data-ing and un-data-ing" (Strohmayer & Muller, 2023), "Data silences: How to unsilence the uncertainties in data science" (Muller, 2024), and a draft under revision on data parables (Muller and Morrison). On this page, I provide background thoughts and then one or more haiku for selected aspects of data work. Thereby, the combination of background and haiku becomes a series of haibuns. The haiku follow the contemporary tradition of English-language haiku, which relaxes requirements for specific syllable counts, imagery, kigo (season words), or kireji (cutting words).
Data are mediated representations - translations - from an original context into an analytic context.
data become the words we use
the words we choose to write
to write the data story
Many papers, books, and commercial offerings recommend storytelling with data. A second version of the haiku emphasizes human choice and construction in such data-stories.
data become the words we use
the words we choose when
making the data story
As Bowker observed, 'raw' data do not exist. All data are mediated through human data attention (human choices and decisions). Katelyn Morrison and I wrote a series of parables that pose questions and challenges for critical data studies and data attention (under revision). The following haiku implies a unstated context [note 1, below]. There is always a context for data, but we may or may not preceive it, and there is a possibility that we may never fully know it.
However this
could also be
a parable
(variant:
However this poem
could also be
a parable)
What does this
parable
refer to
Joni Seagers defined a series of inquiries in feminist and feminicide data studies, in which she remarked that, "What gets counted, counts."
Who gets counted
Who counts whom on behalf of whom
Who said you could count me
Mimi Onuoha created a Library of Missing Datasets, which she described as data silences. Her work became the basis for our papers on "forgetting practices" and "data silences." We learned of her work in Catherine D'Ignazio's and Lauren Klein's Data Feminism, which encouraged us to question any binary proposition as being both too simple and too likely to serve the needs of people in power, while harming people with less power.
(The first of these haikus re-uses a haiku structure by Marlene Mountain [2].)
The data inside the data
The data beneath the data
What exists between the binary extremes
What exists beyond the binary extremes
When we communicate about data and data-distortions, our words may mean one thing to the writer, and another thing to the reader. We connect, but through indirections and resistances.
Between you
and me
a semicolon
Between people
and data
a semicolon
We connect partially. If we mistake our meanings, we may cause harms. Data silences impose subtle or abrupt discontinuities in our connections, and risk causing harms to others or to ourselves.
Between the human
and their data
the silence
And when we think in bigger, more quantitative terms, we may magnify those harms. Melissa Fleming described statistics as "human beings with the tears dried off." Therefore
Between the human
and their data
their tears
Whose tears are on the data
Whose tears become the data
Will you dry the tears
Notes
[1] Mary Oliver used a similar way to ask people to consider what they do not already know.
[2] A similar, less deliberately poetic structure has been independently conceived by Bruce Davis; Tovi Scruggs-Hussein; and Berk Özler.