The Doors Reimagined

A revisit of an earlier notebook on the music band The Doors.

Nearly ten years ago, I put together a small notebook about The Doors. It had lyric fragments, a Jim-shaped word cloud, a loose scatter of recurring themes, and that early sense that data analysis could feel atmospheric, not just explanatory.

This notebook goes back to that experiment. The point is not just to update the code.

It is to revisit the same material from a different angle, with better tools, stronger visuals, more room to explore, and a clearer structure, without losing what made the original interesting in the first place.

At its core, this is still a glance at lyrics, repetition, performance, and the themes that run through The Doors’ work.

Setup complete.
Base directory : /Users/omid/Documents/The-Doors-Reimagined
CSV path       : /Users/omid/Documents/The-Doors-Reimagined/doors_from_kaggle.csv
Mask path      : /Users/omid/Documents/The-Doors-Reimagined/images/jim.jpg
CSV exists     : True
Mask exists    : True

Act I: Open the Doors

Before symbols start clustering and motifs start recurring, it helps to ask a simpler question:

What exactly is in this corpus?

This first act is not interpretive yet. It is forensic. We load the songs, clean the text just enough to make it usable, and take a first look at its physical shape: how many songs there are, how long they run on the page, where the outliers are, and whether the archive looks intact.

Before myth comes inventory.

Corpus summary
- rows: 97
- artists: 1
- unique_songs: 97
- duplicate_song_titles: 0
- duplicate_full_rows: 0
- empty_text_rows: 0
- min_word_count: 78
- median_word_count: 154.0
- mean_word_count: 179.02061855670104
- max_word_count: 753

Null summary
artist    0
song      0
text      0
dtype: int64

First five songs

	artist	song	word_count_rough	char_count
0	Doors	American Night	101	551
1	Doors	Angels And Sailors	250	1372
2	Doors	Been Down So Long	150	733
3	Doors	Black Train Song (live)	329	1717
4	Doors	Build Me A Woman	109	544

Shortest songs by rough word count

	song	word_count_rough	char_count
0	Spanish Caravan	78	385
1	Lions In The Street	82	428
2	Unhappy Girl	83	418
3	Go Insane	83	420
4	Strange Days	90	479
5	Dead Cats, Dead Rats	93	481
6	Wild Child	98	518
7	Summer's Almost Gone	99	541
8	Newborn Awakening	99	531
9	I Can't See Your Face	99	482

Longest songs by rough word count

	song	word_count_rough	char_count
0	Rock Is Dead	753	3782
1	The End-Live In Florida	566	2872
2	The Soft Parade	518	2772
3	Heroin	364	1697
4	The End	342	1694
5	Black Train Song (live)	329	1717
6	When The Music's Over	324	1663
7	Gloria	308	1594
8	Mental Floss	298	1616
9	Shaman's Blues	278	1414

“Spanish Caravan” appears here as the shortest song in the corpus, an underrated gem. It is a brief one, but does not feel small.

Carry me, caravan, take me away
Take me to Portugal, take me to Spain
Andalucia with fields full of grain
I have to see you again and again
Take me, Spanish caravan
Yes, I know you can
Trade winds find Galleons lost in the sea
I know where treasure is waiting for me
Silver and gold in the mountains of Spain
I have to see you again and again
Take me, Spanish caravan
Yes, I know you can

The longest song is called “Rock Is Dead”. It runs slightly longer than an hour and is more of an extended improvisational jam than a fully structured, finished song. It meanders, feels rough around the edges, and the sound quality is not the best, but it is still a fun listen, especially in the background or on a road trip.

Act II: Preparing the Lexicon

Before we can map themes, measure recurrence, or compare songs to one another, the language has to be prepared.

Lyrics are unruly things. They repeat, stutter, and sometime break grammar on purpose. So the goal here is not to sterilize the text, but to make it usable while keeping its pulse intact.

This pass creates a working lexicon for the notebook: lowercased, lightly cleaned, tokenized, and reduced just enough to reveal the recurring vocabulary.

Lexicon summary
- total cleaned tokens: 7,835
- unique cleaned tokens: 1,937
- median cleaned tokens per song: 70.0
- mean lexical diversity: 0.563

Most common cleaned words

	word	count
0	love	149
1	know	89
2	tell	74
3	night	72
4	time	69
5	see	66
6	want	60
7	little	59
8	end	55
9	long	52
10	back	48
11	take	48
12	people	45
13	girl	44
14	child	44
15	please	42
16	good	40
17	ride	38
18	roll	38
19	need	37
20	together	36
21	friend	34
22	rock	34
23	mind	34
24	blue	34

Songs with the highest lexical diversity

	song	token_count	unique_token_count	lexical_diversity
0	Lions In The Street	42	42	1.000000
1	Severed Garden	86	83	0.965116
2	The Movie	66	63	0.954545
3	Stoned Immaculate	55	51	0.927273
4	Angels And Sailors	135	124	0.918519
5	Orange County Suite	81	73	0.901235
6	Latino Chrome	80	72	0.900000
7	Newborn Awakening	51	44	0.862745
8	The Ghost Song	121	103	0.851240
9	Whiskey,mystics And Men	66	56	0.848485

Songs with the lowest lexical diversity

	song	token_count	unique_token_count	lexical_diversity
0	I Looked At You	53	11	0.207547
1	Alabama Song	94	21	0.223404
2	Do It	83	19	0.228916
3	You're Lost Little Girl	42	10	0.238095
4	End Of The Night	58	14	0.241379
5	Light My Fire	82	21	0.256098
6	The Spy	59	16	0.271186
7	I'm A King Bee	105	29	0.276190
8	Don't Go No Farther	82	23	0.280488
9	People Are Strange	53	16	0.301887

Most repetitive songs

	song	token_count	unique_token_count	repetition_ratio
0	I Looked At You	53	11	0.792453
1	Alabama Song	94	21	0.776596
2	Do It	83	19	0.771084
3	You're Lost Little Girl	42	10	0.761905
4	End Of The Night	58	14	0.758621
5	Light My Fire	82	21	0.743902
6	The Spy	59	16	0.728814
7	I'm A King Bee	105	29	0.723810
8	Don't Go No Farther	82	23	0.719512
9	People Are Strange	53	16	0.698113

Sample tokenization

	song	tokens
0	American Night	[hail, american, night, know, sound, gun, thunder, listen, listen, listen, listen, know, many, people, believe, astrology, sagittarius, philosophical, sign, anyway, believe, think, bunch, bullshit...
1	Angels And Sailors	[angel, sailor, rich, girl, backyard, fence, tent, dream, watching, narrowly, soft, luxuriant, car, girl, garage, stripped, liquor, clothes, half, gallon, wine, six, pack, beer, jumped, humped, bo...
2	Been Down So Long	[goddamn, long, look, damn, long, look, people, c'mon, set, free, warden, warden, warden, break, lock, key, warden, warden, warden, break, lock, key, along, mister, c'mon, poor, boy, knee, knee, c...
3	Black Train Song (live)	[recorded, live, spectrum, philadelphia, people, ready, train, coming, need, ticket, clime, board, woooooooooo, train, arrive, sixteen, cortege, long, train, arrive, sixteen, cortege, long, train,...
4	Build Me A Woman	[give, witness, darling, need, witness, babe, wow, poontang, blue, poontang, blue, top, head, bottom, cowboy, shoe, build, woman, make, ten, foot, tall, build, woman, make, ten, foot, tall, make, ...

The lexicon is beginning to show its shape. Certain words rise at once: love, night, time, end, people, girl, child. None of them is surprising on its own, but together they begin to form the small recurring weather of the songs. The language moves between closeness and distance, body and atmosphere, desire and ending. Repetition matters too. Some songs gather force by returning to the same few words until they begin to sound spell-like. Others move more loosely and seem to discover themselves as they go.

“Lions in the Street” stands out in the corpus for its extreme lexical variety. In my cleaned data, every token in the piece is unique and appears only once, which gives it an unusually high lexical diversity, especially for such a short text. At the same time, it should be noted that “Lions in the Street” is not really a freestanding song, but the opening section, or prelude, to “Celebration of the Lizard,” a longer and more theatrical composition. That context matters: rather than building through repetition, the piece moves by rapid image-making, introducing new nouns, settings, and symbolic elements almost line by line. It reads less like a conventional song lyric than like a fragment of ritualized poetry or dream narrative.

By contrast, “I Looked at You” is the most repetitive song in the corpus. It has the lowest lexical diversity, with just 11 unique tokens across 53 total tokens, and that tells you something about how the song works. It is built out of a very small vocabulary, but that vocabulary is used to full effect. The song is kinetic from the start: one person looks, the other looks back, and the whole thing immediately moves forward. It is really about the first spark of attraction and the strange inevitability that can follow from something as small as a glance. There is nothing elaborate about it, and that is part of its strength. It is simple, direct, full of momentum, and one of my favorites. It may not be as well known as some of the Doors’ more famous songs, but I love the energy in it. For all its simplicity, the song captures passion and young love with unusual clarity.

I looked at you
You looked me
I smiled at you
You smiled at me
And we’re on our way
No, we can’t turn back
Yeah, we’re on our way
And we can’t turn back
’Cause it’s too late
Too late, too late
Too late, too late

Act III: The First Constellation

Now that the language has been prepared, the next question is: Which words stay near each other?

A corpus does not only have favourite words. It also has habits of adjacency. Certain terms return in pairs, or keep reappearing in the same local weather. This is one of the first places where a lyric world begins to show itself.

Recurring bigrams across songs

	bigram	count	song_spread
0	little girl	16	6
1	night long	13	5
2	good time	9	5
3	set free	7	5
4	friend end	11	4
5	end night	15	3
6	tell bout	12	3
7	brand new	7	3
8	good old	6	3
9	little darlin	5	3
10	take walk	5	3
11	poor boy	4	3
12	tonight love	4	3
13	around neck	3	3
14	find new	3	3
15	little ride	3	3
16	look eye	3	3
17	take little	3	3
18	take long	3	3
19	whole thing	3	3

What matters here is not only how often a phrase appears, but how widely it travels through the corpus. A repeated phrase in one song may simply belong to a refrain. A phrase that returns across several songs begins to feel more like part of the band’s lyric vocabulary.

“Little girl,” for instance, appears sixteen times across six different songs. In that context it reads less as an isolated phrase than as a recurring form of address, joined by variants like “little darlin.” It suggests tenderness, desire, and a certain idiom of intimacy. Elsewhere the language turns toward night, duration, release, and ending: “night long,” “set free,” “end night.” Taken together, these pairings begin to sketch a world in which terms of endearment sit close to restlessness, darkness, and the thought of an ending still waiting to arrive.

Act IV: The Face in the Lexicon

As before, the word cloud is a simple form, but still a useful one. It lets the dominant words gather into a visible shape.

Here the language returns to the old mask, and the lexicon begins to resemble a face.

Act V: Signature Words

The word cloud gathers the songs into a single surface. This section lets them separate again.

Each song carries a few words that belong to it more strongly than to the rest of the corpus. They are not the whole meaning of the song, and they are not always the most beautiful words on the page. But they do mark its center of gravity.

The aim here is simply to see what rises when each song is allowed to speak a little more distinctly.

	song	signature_terms
0	The Crystal Ship	kiss, another, filled, rather
1	The End	kill, bus, end, west
2	Light My Fire	fire, light, set, try
3	People Are Strange	strange, remembers, face, uneven
4	Moonlight Drive	moonlight, drive, swim, tide
5	Riders On The Storm	storm, rider, thrown, actor
6	Spanish Caravan	caravan, spain, spanish, take

The results are limited to the top few words, but they are already revealing.

In “The Crystal Ship”, the language turns inward. It is one of the tenderest and loveliest songs I have heard from this band, apparently written as a melancholic farewell to Jim’s then girlfriend. In “The End”, the language becomes stark and final. It is one of the most powerful apocalyptic songs I know, forever linked in my mind with Apocalypse Now, even though I had loved the music long before I saw the film. “Light My Fire” builds from a small set of elemental words. “People Are Strange” centers on estrangement and perception. “Moonlight Drive” opens into motion and invitation. “Riders on the Storm”, which happens to be the last song Jim Morrison recorded, darkens into atmosphere and threat, an absolute masterpiece. Even “Spanish Caravan,” brief as it is, carries distance and longing.

Act VI: Song Constellations

Some songs in a corpus seem to call across distance. They may not share a title, a tempo, or a place on an album, but their language begins to lean in the same direction.

This section looks for those affinities. Using the cleaned lyric vocabulary, each song is compared to the others to see which ones live nearest to it in verbal space.

The point is not to force a grand system. It is simply to see which songs begin to cluster when only the language is allowed to decide.

Nearest songs to: The Crystal Ship

	song	similarity
0	Shaman's Blues	0.149
1	Tightrope Ride	0.110
2	The End	0.093
3	I Can't See Your Face	0.089
4	We Could Be So Good Together	0.087

Nearest songs to: The End

	song	similarity
0	The End-Live In Florida	0.403
1	End Of The Night	0.122
2	Crawling King Snake	0.117
3	When The Music's Over	0.115
4	The Ghost Song	0.109

Nearest songs to: Light My Fire

	song	similarity
0	When The Music's Over	0.145
1	I Will Never Be Untrue	0.128
2	End Of The Night	0.121
3	Runnin' Blue	0.119
4	Soul Kitchen	0.110

Nearest songs to: People Are Strange

	song	similarity
0	Strange Days	0.207
1	I Can't See Your Face	0.167
2	L'america	0.110
3	Severed Garden	0.097
4	Hello, I Love You	0.095

Nearest songs to: Moonlight Drive

	song	similarity
0	Moonlight Drive (Including Horse Latitudes)	0.713
1	Not To Touch The Earth	0.086
2	Yes, The River Knows	0.085
3	Take It As It Comes	0.083
4	Ship Of Fools	0.079

Nearest songs to: Riders On The Storm

	song	similarity
0	The Hitchhiker	0.564
1	Who Scared You	0.149
2	End Of The Night	0.138
3	The End	0.090
4	The Soft Parade	0.088

Nearest songs to: Spanish Caravan

	song	similarity
0	You're Lost Little Girl	0.174
1	L'america	0.096
2	Carol	0.087
3	Waiting For The Sun	0.084
4	The End	0.073

These pairings are interesting, but they should be read lightly. Some of them reflect genuine affinities. Others are only the result of local lexical overlap, or the presence of alternate versions of the same song. A better thematic analysis should use semantics-aware representations rather than bag-of-words.

Semantic stack ready.

Act VII: A Semantic Map of the Songs

The earlier map was lexical. It placed songs near one another when they reused similar words.

This one asks for something larger. Instead of comparing surface vocabulary, it compares the songs in a semantic space built from sentence embeddings. That does not eliminate interpretation, but it does let songs remain close even when they do not share the same visible words.

To keep the figure anchored in the lyrics themselves, each song is also given a brief extractive summary: a few lines from the lyric that sit closest to the song’s own semantic center.

Most semantically outlying songs

	song	distance_from_center	representative_lines
0	People Are Strange	0.590426	People are strange when you're a stranger / People are strange when you're a stranger / When you're strange
1	Strange Days	0.530044	Strange days have found us / Strange days have tracked us down / Strange days have found us
2	Dawn's Highway	0.487747	Indians scattered on dawn's highway bleeding / Is that the souls of the ghosts of those dead Indians / Indians scattered on dawn's highway bleeding
3	Lions In The Street	0.464823	Lions in the street and roaming / A beast caged in the heart of a city / With a strange creature groaning beside him
4	My Eyes Have Seen You	0.462690	My eyes have seen you / My eyes have seen you / My eyes have seen you
5	Severed Garden	0.416828	Live in the life of certain, south, cruel bindings / Royal babies, rubies must now replace aborted strangers in the mud / This other kingdom seems by far the best
6	Hour For Magic	0.415893	Resident mockery give us an hour for magic / Give us creed to believe nights of lust / Give of color hundred hues a rich mandala for me and for you
7	Rock Me	0.415455	Rock me all night long / You gotta rock me, little woman / Rock me all night long
8	Mental Floss	0.409035	All deserters will be shot / The assassination occurred at 3.30 / I'm an American! You can't touch me!
9	Get Out Of My Life, Woman	0.397011	Get out of my life, woman / Get out of my life, woman / Get out of my life, woman
10	Whiskey,mystics And Men	0.385748	Well I'll tell you a story of whiskey and mystics and men / And about the believers, and how the whole thing began / Well we'd have no forgiveness forgetfulness faithful remorse
11	Lament	0.384226	Death and my cock are the world / For the death of my cock's spirit / All join now and lament the death of my cock

From the numbers alone, the model identifies “People Are Strange” and “Strange Days” as the two strongest semantic outliers in the corpus. That is plausible enough: both songs are built around a tightly concentrated atmosphere of estrangement, distortion, and uneasy perception, and that kind of thematic consistency can make a song stand apart strongly in an embedding space.

The third major outlier, and the one that appears most visibly isolated on the scatterplot, is “Dawn’s Highway”. What makes it especially interesting is that it stands apart for a different reason. It is not really a conventional song, but a short spoken recollection by Jim Morrison, and its conversational, anecdotal style separates it at once from the more lyrical language of the rest of the catalogue.

In it, Morrison recalls a childhood memory from 1947, when his family came upon a truck accident on a desert highway in New Mexico, with injured people lying in the road. He later said the scene marked him deeply, even claiming that the spirits of the dying people “leaped into his soul”, and he treated the moment as formative for his imagination. The famous line “Indians scattered on dawn’s highway bleeding” later appears in “Peace Frog,” and the same incident echoes through several of his poems. The piece itself stands so far apart semantically because it is less a song than a personal origin story told in Morrison’s own voice.

“My Eyes Have Seen You” also forms an island on our map and appears as an outlier in the visualization. A closer look at the lyrics helps explain why. The style is distinctive: the language is fragmented and highly repetitive, built from short declarative lines that return again and again throughout the song.

If we strip away the repetition and keep only the unique lines, the lyric reduces to a small set of short statements:

My eyes have seen you
Free from disguise
Gazing on a city under television skies
Let them photograph your soul
Memorize your alleys on an endless roll
Meet inside
Show me some more
Stand in the door
Fix your hair
Move upstairs

Taken this way, the song feels less like a narrative than a series of acts of looking. It mixes desire, display, and urban imagery, giving the lyric a voyeuristic and strangely cinematic quality.

Act VIII: Provinces of Theme

A map shows position, but not yet region.

The next question is whether the songs gather into larger semantic neighborhoods: not fixed categories, but loose provinces of feeling, imagery, and address. Some may cluster around estrangement, some around ritual and vision, some around direct desire, some around movement, weather, or night.

What follows is simply an attempt via the BERTopic package to see whether the songs begin to form semantic regions of their own.

OMP: Info #276: omp_set_nested routine deprecated, please use omp_set_max_active_levels instead.

Topic summary

	Topic	Count	Name	topic_label
0	-1	22	-1_scream_along_highway_walkin	Outliers
1	0	13	0_night long_babe_gotta love_heard	night long / babe / gotta love
2	1	10	1_song_sing_listen_rockin	song / sing / listen
3	2	8	2_night night_stay_endless_tell love	night night / stay / endless
4	3	6	3_love take_ride_trip_world	love take / ride / trip
5	4	6	4_cry please_cry_tonight love_take little	cry please / cry / tonight love
6	5	5	5_song_nothin_mercy_ride	song / nothin / mercy
7	6	5	6_words_dawn highway_bleeding ghosts_scattered dawn	words / dawn highway / bleeding ghosts
8	7	5	7_lives_live_little girl_street	lives / live / little girl
9	8	4	8_dances_ghosts_words_dreams	dances / ghosts / words
10	9	4	9_sister lived_killer awoke_tried die_mother	sister lived / killer awoke / tried die
11	10	3	10_hopeless night_goes_wandering_listen	hopeless night / goes / wandering
12	11	3	11_blues_night long_ride_heard	blues / night long / ride
13	12	3	12_grave_wish_freaks_together	grave / wish / freaks

Reading the Provinces

The first BERTopic output is useful, but only provisionally. Its raw labels are mechanical, and at first it leaves a substantial number of songs outside the main thematic regions.

So the next step is to read the provinces more directly, rename them in human terms, and then reassign the outliers to their nearest semantic neighborhood. That gives a more legible map of the corpus without forcing a fixed number of themes in advance.

Original outliers: 22 of 97 songs
Outliers after reduction: 0 of 97 songs

	topic	province	song_count	top_terms	songs
0	0	Barroom Blues and Swagger	20	night long, babe, gotta love, heard, goodbye, listen	Alabama Song | Angels And Sailors | Back Door Man | Break On Through | Cars Hiss By My Window | Crawling King Snake | Don't Go No Farther | Gloria | I'm A King Bee | Land Ho! | Latino Chrome | Mon...
1	1	Direct Address and Pop Desire	10	song, sing, listen, rockin, rise, little darlin	Carol | Do It | Five To One | Get Out Of My Life, Woman | Hello, I Love You | I Looked At You | Love Her Madly | Rock Me | Touch Me | We Could Be So Good Together
2	2	Nocturnal Desire and Estrangement	10	night night, stay, endless, tell love, bliss, night	Easy Ride | I Can't See Your Face | I Will Never Be Untrue | Light My Fire | My Eyes Have Seen You | People Are Strange | Someday Soon | Soul Kitchen | Take It As It Comes | The Crystal Ship
3	4	Moonlit Longing and Tender Distance	8	cry please, cry, tonight love, take little, burn, little ride	Moonlight Drive | Moonlight Drive (Including Horse Latitudes) | Spanish Caravan | Summer's Almost Gone | Waiting For The Sun | Wintertime Love | Wishful Sinful | Yes, The River Knows
4	3	Roads, Rides, and Passage	8	love take, ride, trip, world, fly, born	House Of The Rising Sun | Little Red Rooster | Maggie M'gill | Riders On The Storm | The End | The Hitchhiker | The Soft Parade | Tightrope Ride
5	5	Performance, Protest, and Rough Motion	7	song, nothin, mercy, ride, bring, little darlin	Been Down So Long | Black Train Song (live) | Crossroads | L'america | Runnin' Blue | Tell All The People | Universal Mind
6	9	Death Room, Breakdown, and Shock	6	sister lived, killer awoke, tried die, mother, room sister, nights tried	Go Insane | Heroin | House Announcer | Hyacinth House | Shaman's Blues | The End-Live In Florida
7	7	Girls, Streets, and Intimate Observation	6	lives, live, little girl, street, scene, dream	Love Street | Strange Days | The Spy | Twentieth Century Fox | Unhappy Girl | You're Lost Little Girl
8	8	Spectral Theatre and Dream Speech	6	dances, ghosts, words, dreams, bleeding ghosts, voices	End Of The Night | Hour For Magic | Lions In The Street | Queen Of The Highway | The Ghost Song | Wild Child
9	6	Vision, Origin, and Apocalypse	6	words, dawn highway, bleeding ghosts, scattered dawn, dance, want hear	Dawn's Highway | Lament | Newborn Awakening | Severed Garden | The Movie | When The Music's Over
10	12	Grotesque Public Violence	4	grave, wish, freaks, together, good old, goddamn	Dead Cats, Dead Rats | Mental Floss | Ship Of Fools | The Unknown Soldier
11	10	American Night and Mystic Drift	3	hopeless night, goes, wandering, listen, hopeless, big beat	American Night | Stoned Immaculate | Whiskey,mystics And Men
12	11	Primal Desire and Ritual Blues	3	blues, night long, ride, heard, night, give back	Build Me A Woman | My Wild Love | Woman Is A Devil

	song	topic_final	province
0	American Night	10	American Night and Mystic Drift
1	Stoned Immaculate	10	American Night and Mystic Drift
2	Whiskey,mystics And Men	10	American Night and Mystic Drift
3	Alabama Song	0	Barroom Blues and Swagger
4	Angels And Sailors	0	Barroom Blues and Swagger
...	...	...	...
92	Lament	6	Vision, Origin, and Apocalypse
93	Newborn Awakening	6	Vision, Origin, and Apocalypse
94	Severed Garden	6	Vision, Origin, and Apocalypse
95	The Movie	6	Vision, Origin, and Apocalypse
96	When The Music's Over	6	Vision, Origin, and Apocalypse

97 rows × 3 columns

The raw topic labels are mechanical, but the provinces themselves are readable. Some are stronger than others. A few gather clearly around blues and swagger, direct romantic address, nocturnal estrangement, visionary and apocalyptic speech, or dreamlike theatrical language. Others are smaller and more unstable, more like pockets than full regions.

After reassigning the outliers to their nearest semantic province, the map becomes easier to read as a whole. This does not make the catalogue uniform. It simply makes its larger thematic weather more visible.

	province	song_count
0	Barroom Blues and Swagger	20
1	Direct Address and Pop Desire	10
2	Nocturnal Desire and Estrangement	10
3	Roads, Rides, and Passage	8
4	Moonlit Longing and Tender Distance	8
5	Performance, Protest, and Rough Motion	7
6	Vision, Origin, and Apocalypse	6
7	Spectral Theatre and Dream Speech	6
8	Death Room, Breakdown, and Shock	6
9	Girls, Streets, and Intimate Observation	6
10	Grotesque Public Violence	4
11	American Night and Mystic Drift	3
12	Primal Desire and Ritual Blues	3

The provinces are easier to trust once they can be seen. Some are large and central, others smaller and more singular. The map does not settle interpretation, but it does make the field legible. Songs that seemed isolated in earlier sections can now be placed within a broader semantic neighborhood, while the stranger pieces still keep some of their distance.

Act IX: Emotional Weather

So far, the notebook has traced the shape of the corpus across words, songs, and semantic provinces. The next step is more local. Instead of asking how the songs relate to one another, it asks how feeling moves inside them.

This is only a rough instrument, and it should be read that way. Sentiment analysis does not hear music, irony, or performance, and it cannot fully account for Morrison’s tone. But even with those limits, it can still show something useful: where a song stays level, where it darkens, where it brightens, and where it turns sharply.

Song-level emotional summary

	song	n_lines	mean_compound	min_compound	max_compound	final_compound	mean_intensity	dominant_weather
0	People Are Strange	29	-0.238672	-0.6486	0.0000	-0.2023	0.238672	dark
1	The End	47	0.023794	-0.9851	0.8555	0.0000	0.236487	balanced
2	Moonlight Drive	31	0.024032	-0.5719	0.7964	0.0000	0.101381	balanced
3	Riders On The Storm	33	0.075521	-0.6486	0.7506	0.0000	0.187739	balanced
4	When The Music's Over	58	0.096012	-0.4404	0.8061	0.0000	0.179981	balanced
5	The Crystal Ship	16	0.149500	-0.4767	0.6908	-0.2732	0.306312	balanced

Most important emotional turns

	song	brightest_line	brightest_score	darkest_line	darkest_score	sharpest_turn_line	sharpest_turn_size
0	Moonlight Drive	Easy, I love you	0.796	Baby gonna drown tonight	-0.572	Let's swim out tonight, love	0.815
1	People Are Strange	Streets are uneven, when you're down	0.000	Faces look ugly when you're alone	-0.649	Streets are uneven, when you're down	0.649
2	Riders On The Storm	Gotta love your man, yeah	0.751	There's a killer on the road	-0.649	Girl ya gotta love your man	1.114
3	The Crystal Ship	Another flashing chance at bliss	0.691	You'd rather cry, I'd rather fly	-0.477	The time you ran was too insane	0.842
4	The End	The west is the best, the west is the best	0.856	Kill, kill, kill, kill, kill, kill	-0.985	This is the end, beautiful friend	1.781
5	When The Music's Over	Save us!, Jesus!, save us!	0.806	We're gettin' tired of hangin' around	-0.440	Dance on fire as it intends	1.131

Closing Note

This concludes this exploratory notebook. I wanted to return to The Doors after nearly a decade and look again at aspects of the lyrics that had remained only half-seen in the earlier version: their recurring vocabularies, their shifting emotional weather, their strange semantic neighborhoods, and the different ways individual songs gather force. What emerged, at least for me, is not a single unified lyric world but a set of overlapping climates: intimate, theatrical, spectral, romantic, violent, and sometimes impossible to separate cleanly from one another. I hope the notebook has made that world a little more visible, and that it has been as enjoyable to read as it was to revisit.

Source

• Full repository: The Doors Reimagined
• Standalone notebook render: HTML version