On Graphonyms: the Importance of Chart Type Names

The chart above shows the number of inhabitants of 4 countries living in the country’s capital. How would you call this chart type? I’m asking because this chart type has a lot of names:

That’s a lot of names for a chart that is not even a xenographic (although if you add one dimension, you get the 3-dimensional mosaic plot). But marimekko-mekko-mosaic-proportional stacked bars are not the only chart type suffering from multiple names. It is actually not uncommon for charts to have multiple names:

  •  choropleth maps are sometimes called chloropleth maps (but that’s a different story)

The list goes on: check my Twitter-sourced list or hover over the chart names in the Chartmaker Directory

“Big deal,” I hear you say, “some charts have more than one name.” But I do think this is a problem. Imagine you want to make a marimekko chart, but you don’t know how to do it. So you google “marimekko tutorial” and if you are an Excel or Tableau user, you will probably find what you are looking for. But if you want to make one in R or JMP? Than you need to know it is called a mosaic plot in those tools.

“Just include the name of the tool you want to use in the search,” you respond. True. But then what if you just want to look at some examples of marimekkos/mosaics? Or if you want to learn more about them in general and want to find some background information on them? As long as Google doesn’t know that this chart has more than one name (will eventually happen one day, I guess), you will only find part of the information and examples that are out there. The knowledge about and the applications of the chart type are split into separate silos.

This quote by Steve Haroz, who published studies on 2 chart types, proves that it really is a problem:

That made it tough to lookup related material in both the isotype and connected scatterplot projects.

A common vocabulary to advance

In their excellent paper on the history of the scatterplot, Michael Friendly and Daniel Denis write:

Nomenclature—giving names to things and ideas—provides important markers for the adoption and use of innovations.
Alan Smith, Data Visualisation Editor at the Financial Times and one of the authors of the Visual Vocabulary, echoed the importance of naming charts and sharing the same name for chart types in a mail conversation I had with him:
Soon after arriving at the FT, I quickly thought that having a common (and very public) ‘vocabulary’ was an important thing to do in a newsroom, because it’s a good way of educating both our readers and also the wider newsroom into accepting a wider range of charts than they would otherwise feel comfortable with. Martin Stabe spotted a good example of the impact of this recently where somebody (non-graphics) in editorial conference recently suggested we could ‘do a Sankey’ of a story; that’s not a word that anyone would have recognised in that environment 12 months ago, so proof that we can shift the boundaries.

So in order to popularise common but, more importantly, less common chart types, we should give them good names and we should try to avoid to have “chart type synonyms”.

Here are some things we can do.

1. Name your innovative visualisation

When you come up with new chart type, give it a name. “Obviously,” you say. But nameless chart types are not uncommon. The introductory article of the Temporal Cartogram carries the title “A new kind of map: it’s time“, but doesn’t mention a name for these kind of maps.

The New York times also published these medal ranking methods charts below without a name. News articles are probably not the best place to baptise a new chart type, but I wanted to include them in the xeno.graphics collection, so I asked around. Frontier plot was the one I thought to fit the most.

On other occasions, The New York Times gave their new chart type a name in a making-of article, like was the case for the stacked bump chart (read the making-of).

2. Do the research

I suspect that a lot of chart name synonyms are the result of reinventions. The field is still young and not very well documented, so “inventions” of chart types that someone else already came up with are common. So when you produce some innovative data visualisation on your screen, it is a good idea to do some research to try to find some examples of the same or similar visualisation. If you don’t, you might falsely claim to have invented something and you might add another chart name to the already overcrowded space of graphonyms.

3. Harmonize the catalogues

Xeno.graphics was in part inspired by the great data visualisation catalogues that are out there today:

I haven’t compared the chart names these catalogues use, but they unevitably will use different names for the same chart type. It would be great if all of them would share the same name for the same chart.