A. H. Tammsaare

‘There it is, that Vargamäe,’ the man says, pointing across the bog at the next rise. That is how the book begins — not with a description but with a gesture. Two people are driving uphill, the sun is near the horizon, and the man tells the woman where they are going to live now.

The place exists. It is the Tammsaare-Põhja farm in Vetepere village, Albu parish, Järva county, and it is still there: in the house the writer’s brother built there has been a museum since 1958. But the name Vargamäe stands in no parish register, no manorial record and no map before that cart.

The invented name has beaten the real address so thoroughly that the museum itself says Vargamäe, the road signs say Vargamäe, and so do this page’s own chapter titles. In 2019 Postimees had to run a fact-check headed ‘It really existed, this Vargamäe!’ — not to check the novel, but to check the landscape.

Out of that comes this page’s rule. Every time something is said here, it is also said which of the two is meant: the invention or the real address, the character or the person, the novel or the life. Where two sources part, both stand side by side; the page does not average them.

And so that this does not stay a promise, everything starts with counting. The novel’s first chapter is 79 paragraphs and 2 125 words, of which 876 stand in speech paragraphs — and those figures are not typed here but counted out of the same text this page ships: Vikitekstid’s chapter page at revision 22781, as the pages of its scan read on 18 September 2026.

One thing this page does not know: where Vargamäe is. Wikidata holds no item with a coordinate for the farm or for its museum. So these chapters’ stamps print Vetepere’s coordinate with the village-precision note, and the only sourced point at Vargamäe is the Järva monument, drawn on the map at its computed distance.

Vikitekstid: Tõde ja õigus I/I (rev 22781) · The Vargamäe museum’s historical material · The Postimees fact-check (Siiri Kvell, Kristi Kirss, Kaire Sinimaa of the Vargamäe museum) · Museum Card: the birthplace

Tõde ja õigus appeared in five volumes: 1926, 1929, 1931, 1932 and 1933. Together they hold 190 chapters in Roman numerals, 17 468 paragraphs and — by this page’s own rule — 698 032 running words. The second volume is the largest at 164 037 words; the third the smallest at 123 131.

The text on this page is not an edition but Vikitekstid’s transcription, and every chapter here carries its own revision id. That id belongs to Vikitekstid’s chapter page, which draws its text from the pages of a scan; what is pinned exactly is the text as it was fetched on 18 September 2026 and ships with this page. The scan of volume I is the 1974 Eesti Raamat edition, not the 1926 first edition; the scans of volumes II, III and IV are the Noor-Eesti printings of 1929, 1931 and 1932; the scan of volume V is a 1936 Noor-Eesti printing, not the 1933 first edition. The Literary Museum’s scans of volume III are of the 1931 Tartu edition too. Source and edition are on the shelf card.

One book has three word totals here. This page counts 698 032; the transcription’s own manifest counts 698 040; the research statistics file counts 698 031. The difference is not in the text but in the rule: whether a bare numeral is a word, and whether a foreign letter breaks one. Eight words in seven hundred thousand change nothing except one thing — they remind you that a number about a book is a number about a method.

The whole of the prose is 4.82 megabytes. It is never fetched here in one piece on the page’s own initiative: the shelf below tells you in advance what each volume costs, and only when you press does anything travel. The same holds for the reading clock — it does not run until you start it, and it stops when you stop it.

And then there is a date without which this page could not exist. Estonian copyright runs for the author’s life and seventy years, counted from 1 January of the year after the death. 1940 plus seventy plus one makes 1 January 2011. Seventy full years from the death — and after that anybody may download this text, count it and put it on show.

Vikitekstid: the author index · Vikitekstid: the index of the volume I scan (Eesti Raamat, 1974) · Wikimedia Estonia on copyright and the public domain · The 1926 first edition in the antiquarian trade: 605 or 606 pages

„Tee tööd ja näe vaeva, siis tuleb armastus,“ said the father. So it stands in volume I, chapter XXXIX, paragraph 214, revision 23147. In five volumes, in 698 032 running words, the sentence stands twice: once as spoken and once quoted back. With the word ka it stands not once, although ka alone stands 4 395 times in the novel.

Before it comes a question. In paragraph 213 the son asks why it has to be them who bury their lives in the bog when bread comes easier elsewhere — and adds, thinking it over, ‘if there were love in it, then of course…’ The sentence is the answer. And immediately after it, in paragraph 215, the answer is itself answered: ‘…and love did not come, there is none at Vargamäe to this day.’

Who says it and to whom the research pool could not establish, and the pool’s own refutation warned that ‘Andres to his son Andres’ smelled like a search-engine invention, on the reasoning that Andres’s son is Indrek. The text answers that too. Paragraph 216 names vana Andres and noor Andres in one sentence, riding the cart down from Vargamäe, and three volumes later, at IV/II paragraph 58, the narrator quotes the sentence back as ‘the words of Vargamäe Andres, which that man said to his eldest son when he sent him off to the army, with the taste of death in his own mouth’.

A third time the same construction sounds in another mouth. In volume III, chapter XXVI, a widow in Tallinn says over the money she is lending: ‘He always said love needs fear. You must be afraid, and then love will come — those were his words.’ The same shape, another word where the work stood — and that promise does not come true either.

The word ka is not a small thing. Without it the sentence is about what the work is for; with it, it is a promise that love comes as a bonus on top. The literary scholar Maarja Vaino, as ERR reports her, counts it among the most misunderstood sentences in Estonian cultural history — and ERR’s report of her remark itself prints the sentence with the word ka. So this page stands before one question it can answer only by counting: that word is not in the book.

Vikitekstid: I/XXXIX (rev 23147) · ERR: Tammsaare misinterpreted · A quotation collection, the variant with ka

This text holds 52 577 distinct word forms, and 28 613 of them — more than half, 54.4 per cent — stand exactly once. That is ordinary in a long prose text, and it is also a warning: half the vocabulary is of a kind about which a single occurrence says nothing.

This page has no Estonian morphology and does not pretend to. It counts forms, not lemmas. Which means it cannot tell you how often Tammsaare wrote about love — only how often he wrote armastus, armastust, armastuse and the rest, one form at a time.

That decision is the one handed to you. Type soo and the page shows what it found: soo, soos, sood, soode, soosillal — and also sootuks, sooja, soovi, which are not the bog and have nothing to do with it. You tick which forms are one word. That tick is exactly the point where Estonian begins and a search box ends.

Some families the book itself proposes: armasta- with 35 forms and 797 uses; töö- with 117 forms and 718 uses; kraav- with 69 forms and 405 uses; jumal- with 94 forms and 1 522 uses; soo- with 221 forms and 1 015 uses, false friends included and shown.

If a word is nowhere, that is an answer too and it is printed as one: nowhere in 698 032 words. And one name has to be said here: in 1978 Astrid Villup compiled a frequency dictionary of the author’s and the characters’ speech in volume I. This instrument has a predecessor, it is on paper, and not one figure here has been compared with it.

Maria Mandri, Methis 12/15 · The Frequency Dictionary of Estonian, Tartu

The first rule is the paragraph. In this transcription every speech paragraph opens with „ — 12 231 of the 17 468 — and not one opens with a dash or a guillemet. So the rule is one character, not a decision with three clauses, and the page says so. By that rule 51.4 per cent of all the words are speech.

The second rule is the mark itself: only the words standing between „ and “. That gives 293 640 words, or 42.1 per cent. The two rules are nine and three tenths of a point apart, and the gap is not an error. The gap is the narrator saying who spoke.

The gap comes apart. Ütles stands 3 291 times, vastas 1 834, küsis 1 644, hüüdis 738, rääkis 688, kostis 78, sõnas 8. Around them stand the names — Andres, Indrek, Pearu, Karin, Maurus — and every time the narrator names one, the paragraph rule’s figure grows and the marks rule’s does not.

Volume by volume the difference is wide. Volume I is 35.6 per cent speech, volume II 60.7, volume III 51.1, volume IV 57.4 and volume V 53.6. The first is the volume of bog and ditch; the second is the volume of the school, where Maurus contradicts himself at length and Indrek quarrels with God. The seventy thousand words between the two rules are largely that volume’s.

The longest sentence of the pentalogy comes from there too: volume II, chapter XXVII, paragraph 1, 189 words in one sentence, out of Mr Maurus’s mouth. It is printed here whole and as one paragraph, with its word count in the margin — a sentence that is a picture of itself.

Vikitekstid: II/XXVII (rev 19460) · Dialect in the characters’ speech, Methis 12/15

In all, Andres stands 1 467 times in the five volumes and Indrek 3 981. But that is not the story. The story is where they pass each other: the cumulative totals cross in chapter seventy-eight of a hundred and ninety — volume II, chapter XXXIX — at 1 259 Indreks against 1 242 Andreses. After that the two lines never meet again.

Volume by volume: Indrek 328 / 1 005 / 960 / 901 / 787, Andres 1 224 / 18 / 37 / 0 / 188. The first volume is Andres’s, and in the fourth he is not named once. Pearu, 752 / 11 / 6 / 0 / 265, disappears for three volumes and comes back in the fifth. Karin, 0 / 0 / 0 / 1 198 / 0, exists only in the fourth — and only there is Indrek not the most-named person in the book that is about him.

Maurus, 0 / 544 / 1 / 3 / 0, lives and dies inside the second volume. Krõõt stands 135 times in the first, where she dies, and then 17 times more in the fifth, forty years later. Tiina, 0 / 55 / 0 / 188 / 660, grows slowly and ends the book.

One thing has to be said outright here. These are not people but surface forms. Indrekule and Andresel are not inside these totals, because this page counts forms; and mari stands 754 times, but mari is also a berry that grows in a wood. Every figure here is a figure about a spelling, not about a character, and that must not be forgotten even when the line comes out beautifully.

All five volumes, through the build index

Tammsaare-Põhja was 138.9 hectares, of which 15.6 were arable field. The rest was boggy hay meadow and pasture. That makes 7.9 hectares of other ground for every hectare of field — and in that ratio is all the arithmetic of the first volume. A ditch is dug there because there is little field and a great deal of water.

One of the quarrelled-over ditches is described by the Vargamäe museum’s staff in the Postimees fact-check: the neighbour Jakob Sikenberg’s ditch below the rise was 35 fathoms (sülda) long and 6 wide and stopped at its end with sods, so that the hay meadows could not drain and Peeter Hansen’s animals stood belly-deep in water. The neighbour lowered the dam for the summers only once a court had ruled, and a contract was made of it. The fact-check adds the metres in brackets, about 75 and 13; how long its fathom was it does not say.

The names come from one place: the caption of an 1886 photo-postcard of the home parish names Peeter Hansen as the model for Vargamäe Andres, Jakob Sikenberg as the model for Oru Pearu and Hans Grosstahl as the model for Hundipalu Tiit. The years the first research pass carried — Sikenberg 1869, Hansen 1872 — did not survive checking: 1872 is Hansen’s arrival, 1869 the year the land was surveyed.

And then the word itself. Kraav- gives 69 forms and 405 uses across the five volumes — fewer than you might expect of a book remembered for ditch-digging. Because the ditch is not a subject in this book but an action: it is dug, not discussed.

The court protocols exist and the museum points to them. The real Andres and the real Pearu really did go to law. What became of them in the book is another matter — but the cart they rode in together is in both.

The Vargamäe museum’s historical material · The Vargamäe museum’s visitor guide (the court protocols, the one cart) · Siiri Kvell, Kristi Kirss, Kaire Sinimaa (the Vargamäe museum): the Postimees fact-check, 22 February 2019 · ERR: Jaak Urmet, author of ‘Andres, Pearu, Vargamäe’, on the real Andres and the real Pearu sharing one cart · The 1886 photo-postcard and its caption

Wikidata carries only the new date: 30 January 1878, precision 11, Gregorian. The old-calendar date is not in Wikidata; it comes from the Estonian sources and stands on the timeline as a row of its own. The difference between the two calendars is not looked up: floor(1878/100) − floor(1878/400) − 2 gives twelve, and the build stops if the two birth rows are not exactly twelve days apart. So there are two lives: 22 675 days and 22 687 days.

The schooling runs like this: the village school at Sääsküla, then Prümli, then the Väike-Maarja parish school 1892–1897, then the Hugo Treffner Gymnasium in Tartu 1898–1903, in at twenty and out at twenty-five. That order rests on one summary site the source criticism ruled unusable, so only the years that stand elsewhere are printed here — and the gaps card counts the rest.

From 1903 he works in the office of Konstantin Päts’s paper Teataja. One set of sources ends those years in 1905, another in 1907; both rows are on the timeline. In 1907 he enters the law faculty at Tartu, and in 1911 tuberculosis stops his finals. He took no degree: eight years of study and no degrees at all, and that zero is counted here from an empty array, not typed as a zero.

Then come the years at the end of which there is still no volume. 1912–1913 in the Caucasus at an Estonian settlement — the refutation names Punane Lageda; no source names Sochi, and the map carries two candidate points and two route totals for it. Then six years at his brother Jüri’s at Koitjärve, a village the 1952 training ground emptied. He reaches Tallinn on 1 September 1919; he marries Käthe-Amalie Veltman in March 1920, on the thirteenth or the fourteenth.

From his forty-eighth year to his fifty-fifth, five volumes appear: at 48, 51, 53, 54, 55. Those ages are not typed but computed as the difference of calendar years, because a volume row carries only a year — and the page says which rule it used. Eight years of illness and farm work in front, six years of the silent era behind: 12 March 1934 to 21 June 1940, which outlives him by a hundred and twelve days.

And the name. A. H. Tammsaare is the name of a farm. Whether he ever bore it in law is not established here; Wikidata carries no pseudonym statement at all, and the writers’ museums’ page calls Anton Hansen his civil name. The page says ‘pen name’ and says no more than that.

Wikidata Q353637 (revision 2526924744) · The Estonian Literary Museum’s dated chronology · The Estonian writers’ museums’ life

He lived at that address from 1932 to his death: 2 982 days, computed from two rows of the timeline. The row for the move carries only a year, so the figure counts from 1 January 1932 — and the gaps card counts how many rows on this timeline are read that way. Tõde ja õigus V, Elu ja armastus, Ma armastasin sakslast, Kuningal on külm and Põrgupõhja uus Vanapagan were finished here.

The certificate says a heart attack. The funeral was on 5 March, four days later, and the day was made a national day of mourning. Eesti Kultuurfilm’s newsreel no. 92 filmed it. Nobody on this page has watched that reel, and not one sourced line about the ceremony itself stands here — it is one of this page’s larger holes and it is written on the gaps card.

The last journey is 5.65 kilometres: Kadriorg to Metsakalmistu, computed between two coordinates. Three hundred and fifty-five times shorter than one leg of his life — the Caucasus to Koitjärve, 2 002 kilometres. Drawn at the same scale the last one disappears; which is exactly why it is drawn there.

Metsakalmistu was laid out in 1933 and opened in 1939, a year before his burial. It is kept to look like natural forest: crosses and fences are restricted and a gravestone may be no larger than 80 by 50 centimetres. The grave is monument no. 1114 in the national register and the cemetery itself no. 1088. Estonia’s greatest novelist therefore has one of the smallest possible stones — and that is not an explanation but a rule with a source of its own.

Three and a half months later the state that mourned him was gone. The flat became a museum on his hundredth birthday, 30 January 1978; in 2000 the museum lost rooms to the Kadriorg library; it is shut for renovation now, with reopening promised for the end of 2027.

Tallinn Literary Centre: the Tammsaare museum · The Estonian Wikipedia: A. H. Tammsaare (revision 7076244) — the death and the funeral · The monuments register: grave no. 1114 · Metsakalmistu: a forest that refuses gravestones