Week 4 Session: Geography and habitat
Three records fill the habitat field the same way and mean three different things by it — a mountain at a derivable position on a route, an unspecified sea, and a country placed relative to the entry before it.
15 March 2027 · Phase 1, Build a system that works
The rule this week states
The 山经 routes can be mapped and the 海经 cannot, so a habitat field that accepts both is recording two different kinds of fact under one name.
What this week does#
Week 3 left you a vocabulary of parts you can count. The habitat column has been on the card since week 2 and has never been examined, and it looks as solid as the parts do. It is not. One field name is holding three kinds of claim that cannot be compared with each other.
Three records, each filling that field a different way:
- A position you can derive — 夫诸, filed at a named locus on a named route.
- A field that is filled and locates nothing — 陵鱼, whose habitat is three characters long.
- A position that exists only inside the text — 讙头, placed with respect to whatever was described before it.
A fourth section puts 驺吾 beside those three and asks what one column is supposed to do with all of them.
Example 1: a position you can derive#
The 山經 chapters give, for each mountain, a direction and a distance from the one before. A chain of bearings is a thing you can draw, and the drawing has a length.
Specimen
夫诸
Fuzhu quadricornutum binomial assigned by this course, not by the text
- Kind
- 獸, the word the record uses
- Form
- like a white deer with four horns
- Habitat
- 敖岸之山, first mountain of the 萯山 route
- Omen
- 見則其邑大水 — where it appears, that town floods
- Record
中山經 · 中次三經 · 敖岸之山Central Mountains · third central section · Mount Aoan
有獸焉，其狀如白鹿而四角，名曰夫諸，見則其邑大水。
There is a beast there, shaped like a white deer with four horns. It is called 夫諸. Where it appears, that town floods.
Three clauses and nothing else. The shortest complete record here, which is why a reconstruction of it is so visibly a reconstruction.
Quoted from 山海經 (the Classic of Mountains and Seas), 四庫全書本，郭璞注 — the Siku Quanshu text with 郭璞's (Guō Pú) commentary. Interlinear notes removed; the base text is otherwise unaltered.
夫诸 (fūzhū) is filed under 敖岸之山 on the 萯山 route of 中次三經. That is a named locus, on a named route, at a known place in a sequence. Three facts of the same kind, each checkable against the entries either side of it.
A drawing that is wrong is still the right kind of thing
Whether the route corresponds to terrain anybody has walked is a separate question and not this week’s. The property that matters is that a drawing exists to be wrong about. The next two records do not have that property, and they do not have it in two different ways.
Example 2: a field that is filled and locates nothing#
陵鱼 (língyú) has a record short enough to hold in view whole.
Specimen
陵鱼
Lingyu manipedatum binomial assigned by this course, not by the text
- Kind
- the record never says
- Form
- human-faced, with hands and feet, and a fish's body
- Habitat
- 在海中 — in the sea, and the record says no more
- Record
海內東經Within the Seas, East
陵魚人面，手足，魚身，在海中。
陵魚 has a human face, hands and feet, and a fish's body. It is in the sea.
A habitat that cannot be mapped. The same line stands in 海內北經 only inside 袁珂's note arguing nine passages were misfiled, so even the chapter is contested.
Quoted from 山海經 (the Classic of Mountains and Seas), 四庫全書本，郭璞注 — the Siku Quanshu text with 郭璞's (Guō Pú) commentary. Interlinear notes removed; the base text is otherwise unaltered.
Four statements about a body and one about the world. 在海中: in the sea. Not which sea, not measured from anywhere. The cell is full. Nothing can be derived from what is in it, and a reader who wants to know where to stand has been given the same answer as a reader who wants to know nothing.
The chapter is contested as well. The line stands in 海內北經 only inside 袁珂’s note arguing that nine passages were misfiled, so for this record neither the habitat nor the 卷 can be stated without an argument attached to it.
Example 3: a position that exists only inside the text#
讙头 (huāntóu) fails a third way, and this one is worth more than the other two.
Specimen
讙头
Huantou rostratum binomial assigned by this course, not by the text
- Kind
- the record never says
- Form
- human-faced, winged and beaked; it fishes, and it walks leaning on its wings
- Habitat
- 讙頭國 in the outer south; and the great wilds of the south, spelled 驩頭
- Records
海外南經Beyond the Seas, South
讙頭國在其南，其為人人面有翼，鳥喙，方捕魚。
讙頭國 is to the south of it. Its people have human faces, wings and birds' bills, and they are catching fish.
大荒南經Great Wilds, South
驩頭人面鳥喙，有翼，食海中魚，杖翼而行。
驩頭 has a human face and a bird's bill, and it has wings. It eats fish from the sea, and it walks propped on its wings.
A country in one record and a creature in the other, under two spellings of one name. No field in the entry can hold a thing which is also a people.
Quoted from 山海經 (the Classic of Mountains and Seas), 四庫全書本，郭璞注 — the Siku Quanshu text with 郭璞's (Guō Pú) commentary. Interlinear notes removed; the base text is otherwise unaltered.
The locus is a country, and the country is placed relative to the previous entry: 在其南, to its south. The origin of the coordinate system is the sequence of entries. That makes the position real — you can say exactly where 讙頭國 is, as long as you are reading — and unusable the moment the entry is lifted out of its chapter, which is what filing it on a card does.
In 大荒南經 the same creature appears spelled 驩頭 with no country at all. One of the two records supplies a habitat that depends on its neighbours, and the other supplies none.
What the habitat field is holding#
驺吾 (zōuwú) lands in the same column.
Specimen
驺吾
Zouwu quinquecolorum binomial assigned by this course, not by the text
- Kind
- 獸, the word the record uses
- Form
- the size of a tiger, bearing all five colours, its tail longer than its body
- Habitat
- 林氏國, in the inner north
- Use
- 乘之日行千里 — ride it a thousand 里 in a day
- Record
海內北經Within the Seas, North
林氏國有珍獸，大若虎，五采畢具，尾長于身，名曰騶吾，乘之日行千里。
In 林氏國 there is a treasure-beast, as big as a tiger, bearing all five colours, its tail longer than its body. It is called 騶吾. Ride it a thousand 里 in a day.
Called a 珍獸, a treasure-beast, which is a value and not a kind. The only entry whose most specific classification is a price.
Quoted from 山海經 (the Classic of Mountains and Seas), 四庫全書本，郭璞注 — the Siku Quanshu text with 郭璞's (Guō Pú) commentary. Interlinear notes removed; the base text is otherwise unaltered.
林氏國 is a polity, like 讙頭國, and the record places it nowhere further. So four records put four things in one column: a locus on a measured route, an unbounded sea, a country fixed by its neighbour, and a country fixed by nothing.
Sorting the cards by habitat now produces a list that looks ordered and answers no question. That is the cost of a field that accepts a string, because a string accepts anything. The field is not wrong. It is under-specified, and it hides the difference instead of recording it.
What you leave with#
A position on splitting the habitat column, and a reason for it that is not tidiness. Whoever writes the week 7 proposal either splits the field or defends keeping it whole, and this is the session where the evidence for both options was collected.
If you split it, say what the second field is called and what goes in it when a record gives a country. If you keep it whole, say what your sort order means. Week 5 takes the other column that has been on the card unexamined since week 2, the one holding what happens when you see the creature.
The spec
What you bring to the session. Written so a reader can tell whether it has been met without asking you.
- you have drawn one 山經 route from the stated bearings and distances, without consulting a modern map
- you can name the sentence at which your drawing of a 海經 chapter stopped, and say what was missing from it
- you can state your position on splitting the habitat field, ready to defend in the week 7 proposal
Teaching team
- Marisol Quaye — convenor