a while back i believe it was judith gips commented on one of my dreams and she mentioned that she dreamed more vividly when she had her period, and wondered if i did the same. i replied that i did not have periods, being of the wrong gender for such. she then told me that men of her acquaintance are sensitive to the phases of the moon and so have the equivalence of menstrual periods too. i was unaware of myself having any such response, but it occurred to me that since i have been logging my dreams for many years, i should have enough data to look for temporal cycles in my dreaming.

i can report on this data now. i wrote a perl program that processes the files which contain my dream logs. i got started collecting dreams on 14-DEC-1982 08:19 when i sent a colleague an email in which i described a dream from the night before. at that time i rarely dreamed, or at least, did not remember dreaming when i woke. on that morning i did. a few days later i sent email with a longer dream in it. it seemed to me that this was a handy way of keeping dreams in a machine readable format which i could eventually use for an analysis of sorts. the email also would put a detailed time stamp on the data. i began sending myself emails with dream contents whenever i could remember them.

i have now 369 dreams in these files, containing 40995 lines of text. there is about a 9 year gap from 1985 until 1994 in which i have very few dreams. i recall writing dreams during that time, but apparently i lost them at some time, probably in a hard disk crash. i may eventually recover some as i have crates full of old storage disks around, or those missing dreams may be lost forever. the early part of the dream logs were written in emailers on time sharing computers using a terminal that i kept by my bedside. the time stamps for those come at the end of the writing sessions, when i pressed the send button. the later dreams were written in word processors on microcomputers that i kept beside my bed. those time stamps are written as the first data in the file before i start with the dream content. that is the major systematic change in the temporal aspects of the data collection. the last dream in the log files is dated 10 sep 2001 06:43.

here is how my dreaming happens in relation to the phases of the moon.

Moon Phase of Dreams

0 new moon : 6:******
1 :15:***************
2 :14:**************
3 :16:****************
4 waxing crescent :14:**************
5 :12:************
6 :10:**********
7 first quarter : 8:********
8 :15:***************
9 :16:****************
10 :17:*****************
11 waxing gibbous :12:************
12 :11:***********
13 :14:**************
14 : 8:********
15 full moon :15:***************
16 :10:**********
17 :12:************
18 waning gibbous :14:**************
19 : 9:*********
20 :11:***********
21 :15:***************
22 last quarter :16:****************
23 :11:***********
24 :13:*************
25 : 6:******
26 waning crescent :15:***************
27 :11:***********
28 : 6:******
29 new moon :17:*****************

i see no obvious pattern here, hence i conclude that i have no evidence of a lunar or menstrual cycle. the synodic cycle of the moon is 29.530588861 days in length during the two decades of this data collection. my calculations, being simplistic in nature, vary by a few hours compared to references i can find on the internet, but my data is mainly accurate to the day i wrote the dreams, not the hour. the phases of the moon within the cycle are also rounded off in the graph above. the full moon happens on day 14.77, which i have plotted as 15 here. there are more accurate moon phase algorithms which agree to the minute, but since the method of dream recording does not give that level of percision, day by day analysis ought to suffice.

since i built the code to do this level of temporal analysis, it is of no extra difficulty to plot other time cycles, like day of the week.

Week Day of Dreams

0 monday :57:****************************
1 tuesday :63:*******************************
2 wednesday :40:********************
3 thursday :55:**************************
4 friday :54:***************************
5 saturday :38:*******************
6 sunday :62:*******************************

i was surprised that there is no pattern here which favors weekends, as i believe i have more time on those days to record dreams than on days when i have to rush off to class.

Hour of Day for Dream Writing

0 : 5:**
1 : 3:*
2 : 5:**
3 :11:*****
4 :13:******
5 :24:************
6 :56:****************************
7 :30:***************
8 :45:**********************
9 :58:*****************************
10 :52:**************************
11 :23:***********
12 :17:********
13 : 5:**
14 : 2:*
15 : 3:*
16 : 1:
17 : 1:
18 : 1:
19 : 0:
20 : 0:
21 : 2:*
22 : 0:
23 : 2:*

this data is at the edge of resolution for this kind of record keeping. we do not know when during the night the dreams actually happened, but at least i can know about when i was typing them in. i have been aware for some time that my productive dream cycles seem to happen between 06:00 and 09:30 in the morning. i often work until 03:00 sometimes 04:00 which is my natural time for doing my most creative work. i am often amazed at the content of my dreams which are typed in as fast as i can during the worst time of the day for my mental operations. if i do not get them in quickly, they fade quickly my otherwise poor dream memory. often i do not hardly recall what i have written in the morning until i get home from work that evening and have time to actually read them. i sometimes take naps in the late afternoon, but rarely if ever recall a dream from them.

i suppose i should analyze these dream logs by months to see if there are seasonal effects. i doubt it by looking at the sequence of dream titles. the next step i suppose will be to perform a lexical analysis of the dream contents. maybe the really good dreams actually happen in some occult rhythm that i am unaware of.

. stankuli@etherways.com
=== qui non est hodie cras minus aptus erit
| | who not is today, tomorrow less suitable will be
--- -- Ovid _Remedia Amoris_ i 94