# the why

the theme for day 23 of this year’s #30DayMapChallenge was memory. the question is not only how good is my memory, but how accurate is it, and can it be ✨quantified✨?

# the what

my memory of the place where i spent the first decade and a bit of my life; the house and the road it was on. the field it was opposite. the road to the next village. and the shop where i’d buy the newspapers for my parents and probably some sweets. and the playground a bit beyond that. that bit of path where i first rode a bike. those woods – and all the paths therein. and the tree swing. the local pub with the view over the floodplain, and the further away pub without. that stream a friend and i dammed. the phonebox nonna^[1] taught us how to use. that postbox on the telegraph pole by the gate with the stinging nettles. that hill where i fell off my bike; that corner where my sister fell of hers. the houses of the few other children in the village. the river that flooded and the bridge over it. the stile at end of the path through the field that often had cows in it. the really steep hill that i nearly got to the top of without putting a foot down so many times before eventually making it^[2].

most of those points are within a ~2.5 (±1) km radius. and all are within 5 km. and i reckon i can go further.

# the when

at this point in time^[3] i cannot say exactly when i was last there. to find out i’d have to look at my ride logs, which would involve looking at the map, so i can’t do that yet. i cycled through the village at some point during the winter of 2018/19 (or perhaps 19/20). and before that once around early 2017. and i got close^[4] a few dozen times in the decade before that. and perhaps a very small handful of times in the five years after before that.

# the how

i started with a point, my childhood home. that would be the anchor to reality. i would work in a projected coordinate system, and allowed myself the use of a 100 m grid and a scalebar. working in QGIS, i drew roads and paths, as i remember them. slowly creeping outwards. i added points of interest. and, i kept doing this. eventually i had digitized over 100 km^[5] worth of roads and paths. and 61 points of interest. i could have kept going, and i really did want to keep going. but, well, it is probably sensible to stop doing a slightly frivilous thing at some point.

next step was to create a node at each of the 116 path/road intersections, and give these a number. and then, turn on the real map and add nodes at the junctions, giving them their memory-partner’s number. ditto for the points of interest.

now to jointhe two point layers (memory and reality) and calculate some things:

• true distance^[6] from home
• remembered distance from home^[7]
• true bearing from home
• remembered bearing from home

# pairwise

at this point i am in a position to plot something like fig 1, and fig 2, which is all well and good. but, would we get a better idea of one’s memory if we calculated pairwise^[8] distances between each point in reality-land. and then in memory land. and plotted those against one another?^[9]

justification: because i might have made that road too long, the features at the end of it are too far away, but the two points of interest at the end of it, are correct – relative to one another. that then gives us fig 3. nice. the brighest bits lower down make sense.

# output

the final map looks a bit like this (fig 4)

NOTE: the railway and river were not included in any of distance calculations above.

# some observations

• i thought i detected a dependence in my ability to remember distances on whether or not i mostly cycled a road or drove it. but i’m not sure i can back that up. similarly i thought paths that i mostly cycled uphill were coming out longer than they ought to…
• one minor triumph is that the network appears to be correct. that is, to get from here to there, you really do have to turn left at that junction then, take the second right at that one etc…
• a few rememberd junctions didn’t look quite right as, to use the audax vernacular, they are green triangles rather than a straightforward t-junction. there were also a few straight-ons that might be more accurately described as staggereds and vice-versa.
• i think having a grid visible whilst drawing, encouraged me to adhere to north-south, east-west directions more than i perhaps would have naturally. or, maybe when memory says ‘turn right’ that is assumed to be taken as 90°. both. perhaps.

# footnotes