Why do hymnal comparisons vary drastically by direction?

You are here

Haruo's picture

When I compare two hymnals, and am informed that one of them has 106 hymns in common with the other, it seems to me this should mean that the other one also has 106 hymns in common with the first one. But it is almost never so, and sometimes the differences are quite drastic.

For example, comparing various hymnals to HTLG2017, I get these results:

Hymns in common, comparing first hymnal to second vs. comparing second hymnal to first:

HTLG2017 v. CLH2006 106 / 127
HTLG2017 v. GG2013 148 / 154
HTLG2017 v. CoCS2013 19 / 20
HTLG2017 v. CGH2010 189 / 232
HTLG2017 v. BH2008 112 / 122
HTLG2017 v. ELW2006 132 / 143

By this point you might think the HTLG 2017 v. X comparisons are always weaker than the inverse. But no...

HTLG2017 v. UMH 191 / 184
HTLG2017 v. CH1995 156 / 121
HTLG2017 v. HWC1986 180 / 141
HTLG2017 v. GC1994 116 / 69
HTLG2017 v, VU1996 172 / 134
HTLG2017 v. CHWB1996 205 // 173

...and only two hymnals I have tested show an equal number of hymns in common in both directions:
LEVS1993 tied at 19
ICH5 tied at 130

Just doesn't make sense to me...


Comments

The difference usually has to do with a hymnal repeating a hymn with a different tune. For example CLH2006 has two versions of "Come, thou Fount of Every Blessing" #13 and #14, but HTLG2017 has one version, #253. #253 in HTLG2017 is counted twice when compared against CLH2006, but since HTLG2007 has only one version, only the first version from CLH2006 is counted.

... and there's no effort to compare pairings. It makes sense, especially when comparing DNAH data, but I would find the ability to compare pairings (and tunes, for that matter) highly useful at times.