이 문서는 수학에서
리 대수 를 분류하는 벡터의 집합에 관한 것입니다.
식물학 에서 식물의 뿌리들의 구조에 대해서는
뿌리 문서를, 대한민국의 지명에 대해서는
근계리 문서를 참고하십시오.
리 군 이론에서, 근계 (根系, 영어 : root system )는 일련의 기하학적 성질을 만족하는 유한 차원 벡터 의 집합 이다. 근계의 원소인 벡터는 근 (根, 영어 : root )이라고 부른다. 주어진 근계에 대하여 특정 성질을 만족하는 부분집합인 단순근 (單純根, 영어 : simple root )의 집합을 고를 수 있고, 이를 딘킨 도표 (영어 : Dynkin diagram )로 나타내어 분류할 수 있다. 반단순 리 군 에 근계를 대응시킬 수 있으며, 이를 통해 반단순 리 군들을 분류할 수 있다.
예외 리 군 G2 의 근계. α {\displaystyle \alpha } 와 β {\displaystyle \beta } 는 단순근이다. 모든 근계는 기약 근계 (旣約根系, 영어 : irreducible root system )의 합으로 나타낼 수 있다. 기약 근계(의 동형류)는 복소수체 위의 단순 리 대수 (의 동형류)와 일대일로 대응한다.
정의 유한 차원 실수 내적 공간 ( V , ( ⋅ , ⋅ ) ) {\displaystyle (V,(\cdot ,\cdot ))} 속의 부분 집합 Φ ⊆ V {\displaystyle \Phi \subseteq V} 가 다음 다섯 조건들을 모두 만족시킨다면, 근계 라고 한다.
(선형 생성) V = Span R Φ {\displaystyle V=\operatorname {Span} _{\mathbb {R} }\Phi } . 즉, V {\displaystyle V} 의 모든 원소는 Φ {\displaystyle \Phi } 의 원소들의 선형 결합 으로 나타낼 수 있다. (이는 유일하지 않을 수 있다.) (스칼라배의 제한) α ∈ Φ {\displaystyle \alpha \in \Phi } 라면, − α ∈ Φ {\displaystyle -\alpha \in \Phi } 이고, 그 밖의 다른 스칼라배 t Φ {\displaystyle t\Phi } ( t ∈ R {\displaystyle t\in \mathbb {R} } )는 Φ {\displaystyle \Phi } 의 원소가 아니다. (반사에 대한 닫힘) 임의의 α , β ∈ Φ {\displaystyle \alpha ,\beta \in \Phi } 에 대하여, α {\displaystyle \alpha } 에 대하여 수직인 초평면에 대한 β {\displaystyle \beta } 의 반사 β − 2 α ( α , β ) / ( α , α ) {\displaystyle \beta -2\alpha (\alpha ,\beta )/(\alpha ,\alpha )} 도 Φ {\displaystyle \Phi } 의 원소다. 즉, 근들은 다른 근에 대한 반사에 대하여 닫혀 있다. (정수성) 0 ∉ Φ {\displaystyle 0\not \in \Phi } 이며, ∀ α , β ∈ Φ : 2 ( α , β ) / ( α , α ) ∈ Z {\displaystyle \forall \alpha ,\beta \in \Phi \colon 2(\alpha ,\beta )/(\alpha ,\alpha )\in \mathbb {Z} } 유한 집합 이다.근계의 원소는 근 이라고 부른다. 근계의 계수 (階數, 영어 : rank )는 V {\displaystyle V} 의 차원이다.
두 실수 내적 공간 V {\displaystyle V} , V ′ {\displaystyle V'} 및 그 속의 근계 Φ ⊆ V {\displaystyle \Phi \subseteq V} , Φ ′ ⊆ V ′ {\displaystyle \Phi '\subseteq V'} 에 대하여, 만약 f ( Φ ) = Φ ′ {\displaystyle f(\Phi )=\Phi '} 가 되는 전단사 실수 선형 변환 f : V → V ′ {\displaystyle f\colon V\to V'} 이 존재하며, 또한
( f ( α ) , f ( β ) ) V ′ ( f ( α ) , f ( α ) ) V ′ = ( α , β ) V ( α , α ) V ∀ α , β ∈ Φ {\displaystyle {\frac {(f(\alpha ),f(\beta ))_{V'}}{(f(\alpha ),f(\alpha ))_{V'}}}={\frac {(\alpha ,\beta )_{V}}{(\alpha ,\alpha )_{V}}}\qquad \forall \alpha ,\beta \in \Phi } 라면, ( V , Φ ) {\displaystyle (V,\Phi )} 와 ( V ′ , Φ ′ ) {\displaystyle (V',\Phi ')} 를 서로 동형 이라고 한다.
특히, 동형이 등거리 변환 일 필요는 없다. 예를 들어, 항등 함수 ( V , ( − , − ) ) → ( V , 2 ( − , − ) ) {\displaystyle (V,(-,-))\to (V,2(-,-))} 역시 허용된다. 이 때문에, 통상적으로, 근계에서 가장 긴 근의 노름 을 2 {\displaystyle {\sqrt {2}}} 로 놓는다. (이에 따라, 더 짧은 근의 노름은 1 {\displaystyle 1} 또는 2 / 3 {\displaystyle {\sqrt {2/3}}} 이다.)
통상적으로, 다음과 같은 표기를 사용한다.
⟨ α , β ⟩ = 2 ( α , β ) ( α , α ) {\displaystyle \langle \alpha ,\beta \rangle ={\frac {2(\alpha ,\beta )}{(\alpha ,\alpha )}}} (이는 물론 쌍선형 형식 을 이루지 못한다.)
양근과 단순근 근계 Φ {\displaystyle \Phi } 의 양근의 집합 (陽根의 集合, 영어 : set of positive roots ) Φ + ⊂ Φ {\displaystyle \Phi ^{+}\subset \Phi } 는 다음을 만족하는 부분집합이다.
임의의 α ∈ Φ {\displaystyle \alpha \in \Phi } 에 대하여, α ∈ Φ + {\displaystyle \alpha \in \Phi ^{+}} 이거나 − α ∈ Φ + {\displaystyle -\alpha \in \Phi ^{+}} 이지만, { α , − α } ⊂ Φ + {\displaystyle \{\alpha ,-\alpha \}\subset \Phi ^{+}} 는 아니다. α , β ∈ Φ + {\displaystyle \alpha ,\beta \in \Phi ^{+}} 이고, α + β ∈ Φ {\displaystyle \alpha +\beta \in \Phi } 이면 α + β ∈ Φ + {\displaystyle \alpha +\beta \in \Phi ^{+}} 이다.양근의 집합의 원소를 양근 (陽根, 영어 : positive root )이라고 한다. 양근의 집합 Φ + ⊆ Φ {\displaystyle \Phi ^{+}\subseteq \Phi } 이 주어졌을 때, 격자
{ v ∈ V : ∀ α ∈ Φ : α ∨ ( v ) ∈ Z } {\displaystyle \{v\in V\colon \forall \alpha \in \Phi \colon \alpha ^{\vee }(v)\in \mathbb {Z} \}} 위에 다음과 같은 부분 순서 를 줄 수 있다.
u ≤ v ⟺ ∀ α ∈ Φ + : α ∨ ( v − u ) ≥ 0 {\displaystyle u\leq v\iff \forall \alpha \in \Phi ^{+}\colon \alpha ^{\vee }(v-u)\geq 0} 이 구성은 리 대수 의 표현론 에 등장하며, 이 경우 위의 격자는 정수 무게 의 격자에 해당한다.
단순근 어떤 양근의 집합이 주어졌을 때, 단순근 (單純根, 영어 : simple root )은 두 양근의 합으로 나타낼 수 없는 근이다. 단순근들의 집합은 V {\displaystyle V} 의 기저 를 이룬다.
카르탕 행렬 근계 Φ {\displaystyle \Phi } 와 그 위의 순서를 매긴 단순근 의 열 α 1 , … , α r {\displaystyle \alpha _{1},\dots ,\alpha _{r}} 이 주어졌다고 하자. 그렇다면, 이에 대응하는 카르탕 행렬 (영어 : Cartan matrix ) M {\displaystyle M} 은 다음과 같은 r × r {\displaystyle r\times r} 정사각 행렬 이다.
M = ( M i j ) i , j = 1 , … , r {\displaystyle M=(M_{ij})_{i,j=1,\dots ,r}} M i j = 2 ( α i , α j ) ( α i , α i ) {\displaystyle M_{ij}=2{\frac {(\alpha _{i},\alpha _{j})}{(\alpha _{i},\alpha _{i})}}} 정의에 따라, 카르탕 행렬의 대각선 성분의 값은 모두 2이다.
카르탕 행렬이 주어지면, 이에 대응하는 근계 (및 복소수 반단순 리 대수 )를 재구성할 수 있다.
딘킨 도표 기약근계의 딘킨 도표 각 근계 ( V , Φ ) {\displaystyle (V,\Phi )} 에 대하여, 딘킨 도표 (Дынкин圖表, 영어 : Dynkin diagram )라는, 일종의 유향 그래프 를 대응시킬 수 있다. 우선, 임의로 ( V , Φ ) {\displaystyle (V,\Phi )} 의 양근의 집합 Φ + ⊆ Φ {\displaystyle \Phi ^{+}\subseteq \Phi } 를 고르자.
딘킨 도표는 각 단순근에 대응하는 꼭짓점 을 갖는다. 두 꼭짓점 사이에는 0개, 1개, 2개, 또는 3개의 변(邊)이 존재할 수 있다. 변이 2개 또는 3개인 경우, 변은 방향을 가지며, 이 방향은 항상 더 짧은 단순근을 가리킨다. (이 경우 두 단순근의 길이는 항상 다르다.) 두 꼭짓점 사이의 변의 수는 두 단순근 사이의 각도에 대응하며, 다음 표를 따른다. 근 사이 각 (라디안 ) 근 사이 각 (°) 변의 종류 π / 2 {\displaystyle \pi /2} 90° 변 없음 2 π / 3 {\displaystyle 2\pi /3} 120° 하나의 변 3 π / 4 {\displaystyle 3\pi /4} 135° 두 개의 변 + 화살표 5 π / 6 {\displaystyle 5\pi /6} 150° 세 개의 변 + 화살표
딘킨 도표는 단순근의 선택에 관계없이 동일하다.
기약 근계의 딘킨 도표는 연결되어 있다. 딘킨 도표의 연결 성분 분해는 근계의 (기약 근계들로의) 직합 분해와 같다.
성질 정수성 공리에 따라, 두 근 사이의 각은 π / 2 {\displaystyle \pi /2} , π / 3 {\displaystyle \pi /3} , π / 4 {\displaystyle \pi /4} , π / 6 {\displaystyle \pi /6} 또는 이들의 여각이다. 근계의 정수성은 두 근 사이의 각들을 제한한다. 정수성 공리에 따라, 두 근 사이의 각의 코사인 은 정수의 제곱근의 반이어야 한다.
Z ∋ 2 ( α , β ) ( α , α ) ⋅ 2 ( α , β ) ( β , β ) = 4 ( α , β ) 2 | α | 2 | β | 2 = 4 cos 2 ( θ ) = ( 2 cos ( θ ) ) 2 . {\displaystyle \mathbb {Z} \ni 2{\frac {(\alpha ,\beta )}{(\alpha ,\alpha )}}\cdot 2{\frac {(\alpha ,\beta )}{(\beta ,\beta )}}=4{\frac {(\alpha ,\beta )^{2}}{\vert \alpha \vert ^{2}\vert \beta \vert ^{2}}}=4\cos ^{2}(\theta )=(2\cos(\theta ))^{2}.} 2 cos ( θ ) ∈ [ − 2 , 2 ] {\displaystyle 2\cos(\theta )\in [-2,2]} 이므로,
cos θ = 0 , ± 1 2 , ± 2 2 , ± 3 2 , ± 1 {\displaystyle \cos \theta =0,\pm {\tfrac {1}{2}},\pm {\tfrac {\sqrt {2}}{2}},\pm {\tfrac {\sqrt {3}}{2}},\pm 1} 이다. 즉, θ {\displaystyle \theta } 는 90°, 60° 또는 120°, 45° 또는 135°, 30° 또는 150°, 0° 또는 180°이다.
연산
스칼라배 근계 ( V , Φ ) {\displaystyle (V,\Phi )} 및 임의의 실수 t ∈ R ∖ { 0 } {\displaystyle t\in \mathbb {R} \setminus \{0\}} 및 임의의 직교 행렬 M ∈ O ( V ; R ) {\displaystyle M\in \operatorname {O} (V;\mathbb {R} )} 에 대하여, ( V , t M Φ ) {\displaystyle (V,tM\Phi )} 역시 근계를 이루며, 이는 원래 근계 ( V , Φ ) {\displaystyle (V,\Phi )} 와 동형이다.
직합 두 근계 ( V , Φ ) {\displaystyle (V,\Phi )} , ( V ′ , Φ ′ ) {\displaystyle (V',\Phi ')} 가 주어졌을 때, 그 직합 Φ ⊕ Φ ′ {\displaystyle \Phi \oplus \Phi '} 은 다음과 같은 근계이다.
Φ ⊕ Φ ′ = ι ( Φ ) ∪ ι ′ ( Φ ′ ) {\displaystyle \Phi \oplus \Phi '=\iota (\Phi )\cup \iota '(\Phi ')} V → ι V ⊕ V ′ ← ι ′ V ′ {\displaystyle V{\xrightarrow {\iota }}V\oplus V'{\xleftarrow {\iota '}}V'} 여기서 ι {\displaystyle \iota } 와 ι ′ {\displaystyle \iota '} 은 직합 의 정의에 등장하는 표준 포함 사상이다.
기약 근계 (旣約根系, 영어 : irreducible root system )는 두 (자명하지 않은) 근계의 합이 아닌, 자명하지 않은 근계다. 모든 근계는 기약 근계의 합으로 유일하게 나타낼 수 있다.
기약 근계의 근은 모두 길이가 같거나, 길이가 두 가지가 있다. 길이가 두 가지가 있을 경우는 긴 것은 긴 근 (영어 : long root ), 짧은 것은 짧은 근 (영어 : short root )으로 분류한다. (만약 길이가 모두 같다면, 모든 근이 긴 근이다.) 이 경우, 긴 근과 짧은 근의 노름 의 비는 2 {\displaystyle {\sqrt {2}}} 이다. (통상적으로, 긴 근의 노름은 2 {\displaystyle {\sqrt {2}}} 로, 짧은 근의 노름은 (만약 존재한다면) 1 {\displaystyle 1} 로 잡는다.)
쌍대 근계 근계 ( V , Φ ) {\displaystyle (V,\Phi )} 의 쌍대 근계 (雙對根系, 영어 : dual root system )는 다음과 같다.
V ∨ {\displaystyle V^{\vee }} 는 V {\displaystyle V} 의 (대수적) 쌍대 공간 이다. 물론, 내적을 사용하여 표준적인 동형 사상 V → V ∨ {\displaystyle V\to V^{\vee }} 이 존재한다. Φ ∨ = { α ∨ : α ∈ Φ } {\displaystyle \Phi ^{\vee }=\{\alpha ^{\vee }\colon \alpha \in \Phi \}} 임의의 u , v ∈ V {\displaystyle u,v\in V} 에 대하여, u ∨ ∈ V ∗ {\displaystyle u^{\vee }\in V^{*}} , u ∨ ( v ) = ⟨ u , v ⟩ = 2 ( u , v ) / ( u , u ) {\displaystyle u^{\vee }(v)=\langle u,v\rangle =2(u,v)/(u,u)} 그렇다면 ( V ∨ , Φ ∨ ) {\displaystyle (V^{\vee },\Phi ^{\vee })} 역시 근계를 이룬다.
임의의 근계 ( V , Φ ) {\displaystyle (V,\Phi )} 는 그 이중 쌍대 근계 ( V ∨ ∨ , Φ ∨ ∨ ) {\displaystyle (V^{\vee \vee },\Phi ^{\vee \vee })} 와 표준적으로 동형이다.
단순 근계 가운데, B n {\displaystyle B_{n}} 의 쌍대 근계는 C n {\displaystyle C_{n}} 이다. 다른 단순 근계들( A n {\displaystyle A_{n}} , D n {\displaystyle D_{n}} , E 6 , E 7 , E 8 , F 4 , G 2 {\displaystyle E_{6},E_{7},E_{8},F_{4},G_{2}} )은 스스로의 쌍대 근계이다.
분류
기약 근계의 목록 기약 근계는 다음과 같이 분류한다. 고전 근계 (영어 : classical root system )는 네 개의 족 A n {\displaystyle A_{n}} , B n {\displaystyle B_{n}} , C n {\displaystyle C_{n}} , D n {\displaystyle D_{n}} 으로 나뉘고, 나머지로 다섯 개의 예외 근계 (영어 : exceptional root system ) G 2 , F 4 , E 6 , E 7 , E 8 {\displaystyle G_{2},F_{4},E_{6},E_{7},E_{8}} 이 있다. 그 아래첨자는 근계의 계수다. 고전 근계는 고전군 (직교군 , 특수 유니터리 군 , 심플렉틱 군 )의 리 대수 (의 복소화)의 근계이나, 예외 근계는 그렇지 않다. 아래 표에서는 관례를 따라 긴 근의 길이가 2 {\displaystyle {\sqrt {2}}} 가 되도록 정규화하였다.[1]
근계 근의 수 짧은 근 수 긴 근 부분격자의 지표 카르탕 행렬식 바일 군의 크기 콕서터 수 h {\displaystyle h} 이중 콕서터 수 h ∨ {\displaystyle h^{\vee }} 딘킨 도표 콕서터 라벨[2] :43 이중 콕서터 라벨[2] :43 An (n ≥ 1) n (n + 1) n + 1(n + 1)! n + 1 {\displaystyle n+1} ∙ − ∙ − ⋯ − ∙ {\displaystyle \bullet -\bullet -\cdots -\bullet } 1 − 1 − ⋯ − 1 {\displaystyle 1-1-\cdots -1} Bn (n ≥ 2) 2n 2 2n 2 2 2n n ! 2 n {\displaystyle 2n} 2 n − 1 {\displaystyle 2n-1} ∙ − ∙ − ⋯ − ∙ ⇒ ∙ {\displaystyle \bullet -\bullet -\cdots -\bullet \Rightarrow \bullet } 1 − 2 − ⋯ − 2 ⇒ 2 {\displaystyle 1-2-\cdots -2\Rightarrow 2} 1 − 2 − ⋯ − 2 ⇒ 1 {\displaystyle 1-2-\cdots -2\Rightarrow 1} Cn (n ≥ 3) 2n 2 2n (n − 1) 2 2 2n n ! 2 n {\displaystyle 2n} n + 1 {\displaystyle n+1} ∙ − ∙ − ⋯ − ∙ ⇐ ∙ {\displaystyle \bullet -\bullet -\cdots -\bullet \Leftarrow \bullet } 2 − 2 − ⋯ − 2 ⇐ 1 {\displaystyle 2-2-\cdots -2\Leftarrow 1} 1 − 1 − ⋯ − 1 ⇐ 1 {\displaystyle 1-1-\cdots -1\Leftarrow 1} Dn (n ≥ 4) 2n (n − 1) 4 2n − 1 n ! 2 n − 2 {\displaystyle 2n-2} ∙ − ∙ − ⋯ − ∙ < ∙ ∙ {\displaystyle \bullet -\bullet -\cdots -\bullet <{\bullet \atop \bullet }} 1 − 2 − ⋯ − 2 < 1 1 {\displaystyle 1-2-\cdots -2<{1 \atop 1}} E6 72 3 27 ×34 ×5 12 ∙ − ∙ ∙ − ∙ > ∙ − ∙ {\displaystyle {\bullet -\bullet \atop \bullet -\bullet }>\bullet -\bullet } 1 − 2 1 − 2 > 3 − 2 {\displaystyle {1-2 \atop 1-2}>3-2} E7 126 2 210 ×34 ×5×7 18 ∙ − ∙ ∙ > ∙ − ∙ − ∙ − ∙ {\displaystyle {\bullet \atop {}}{- \atop {}}{\bullet \atop \bullet }>\bullet -\bullet -\bullet -\bullet } 2 − 3 2 > 4 − 3 − 2 − 1 {\displaystyle {2 \atop {}}{- \atop {}}{3 \atop 2}>4-3-2-1} E8 240 1 214 ×35 ×52 ×7 30 ∙ − ∙ ∙ > ∙ − ∙ − ∙ − ∙ − ∙ {\displaystyle {\bullet \atop {}}{- \atop {}}{\bullet \atop \bullet }>\bullet -\bullet -\bullet -\bullet -\bullet } 2 − 4 3 > 6 − 5 − 4 − 3 − 2 {\displaystyle {2 \atop {}}{- \atop {}}{4 \atop 3}>6-5-4-3-2} F4 48 24 4 1 27 ×32 12 9 ∙ − ∙ ⇒ ∙ − ∙ {\displaystyle \bullet -\bullet \Rightarrow \bullet -\bullet } 2 − 3 ⇒ 4 − 2 {\displaystyle 2-3\Rightarrow 4-2} 2 − 3 ⇒ 2 − 1 {\displaystyle 2-3\Rightarrow 2-1} G2 12 6 3 1 22 ×3 6 4 ∙ ⇛ ∙ {\displaystyle \bullet \Rrightarrow \bullet } 2 ⇛ 3 {\displaystyle 2\Rrightarrow 3} 2 ⇛ 1 {\displaystyle 2\Rrightarrow 1}
고전적 기약 근계 A n {\displaystyle A_{n}} 형 근계의 단순근은 다음과 같다. (편의상 R n + 1 {\displaystyle \mathbb {R} ^{n+1}} 의 원소로 표기하였다.)
α 1 = ( 1 , − 1 , 0 , … , 0 , 0 ) {\displaystyle \alpha ^{1}=(1,-1,0,\dots ,0,0)} α 2 = ( 0 , 1 , − 1 , … , 0 , 0 ) {\displaystyle \alpha ^{2}=(0,1,-1,\dots ,0,0)} ⋮ {\displaystyle \vdots } α n = ( 0 , 0 , … , 1 , − 1 ) {\displaystyle \alpha ^{n}=(0,0,\dots ,1,-1)} B n {\displaystyle B_{n}} 형 근계의 단순근은 다음과 같다.
α 1 = ( 1 , − 1 , 0 , … , 0 , 0 ) {\displaystyle \alpha ^{1}=(1,-1,0,\dots ,0,0)} α 2 = ( 0 , 1 , − 1 , … , 0 , 0 ) {\displaystyle \alpha ^{2}=(0,1,-1,\dots ,0,0)} ⋮ {\displaystyle \vdots } α n − 1 = ( 0 , 0 , … , 1 , − 1 ) {\displaystyle \alpha ^{n-1}=(0,0,\dots ,1,-1)} α n = ( 0 , 0 , … , 0 , 1 ) {\displaystyle \alpha ^{n}=(0,0,\dots ,0,1)} C n {\displaystyle C_{n}} 형 근계의 단순근은 다음과 같다.
α 1 = ( 1 , − 1 , 0 , … , 0 , 0 ) {\displaystyle \alpha ^{1}=(1,-1,0,\dots ,0,0)} α 2 = ( 0 , 1 , − 1 , … , 0 , 0 ) {\displaystyle \alpha ^{2}=(0,1,-1,\dots ,0,0)} ⋮ {\displaystyle \vdots } α n − 1 = ( 0 , 0 , … , 1 , − 1 ) {\displaystyle \alpha ^{n-1}=(0,0,\dots ,1,-1)} α n = ( 0 , 0 , … , 0 , 2 ) {\displaystyle \alpha ^{n}=(0,0,\dots ,0,2)} D n {\displaystyle D_{n}} 형 근계의 단순근은 다음과 같다.
α 1 = ( 1 , − 1 , 0 , … , 0 , 0 ) {\displaystyle \alpha ^{1}=(1,-1,0,\dots ,0,0)} α 2 = ( 0 , 1 , − 1 , … , 0 , 0 ) {\displaystyle \alpha ^{2}=(0,1,-1,\dots ,0,0)} ⋮ {\displaystyle \vdots } α n − 1 = ( 0 , 0 , … , 1 , − 1 ) {\displaystyle \alpha ^{n-1}=(0,0,\dots ,1,-1)} α n = ( 0 , 0 , … , 1 , 1 ) {\displaystyle \alpha ^{n}=(0,0,\dots ,1,1)}
예외적 기약 근계 예외적 기약 근계는 E₆ , E₇ , E₈ , F₄ , G₂ 총 5개가 있다. 이들의 단순근들은 다음과 같다.
E8 1 -1 0 0 0 0 0 0 0 1 -1 0 0 0 0 0 0 0 1 -1 0 0 0 0 0 0 0 1 -1 0 0 0 0 0 0 0 1 -1 0 0 0 0 0 0 0 1 -1 0 0 0 0 0 0 1 1 0 -½ -½ -½ -½ -½ -½ -½ -½
G2 2 / 3 {\displaystyle {\sqrt {2/3}}} 0 − 3 / 2 {\displaystyle -{\sqrt {3/2}}} 1 / 2 {\displaystyle 1/{\sqrt {2}}}
예
낮은 차원의 근계 0차원 근계는 (자명하게) 하나 밖에 없다.
1차원 근계는 하나 밖에 없으며, { − 2 , 2 } ⊂ R {\displaystyle \{-{\sqrt {2}},{\sqrt {2}}\}\subset \mathbb {R} } 이다.
2차원 근계는 총 4개가 있으며, 이들 가운데 3개는 기약 근계이다. (아래 표에서, B 2 {\displaystyle B_{2}} 와 C 2 {\displaystyle C_{2}} 는 서로 동형이며, A 1 × A 1 {\displaystyle A_{1}\times A_{1}} 과 D 2 {\displaystyle D_{2}} 역시 서로 동형이다.)
A 1 × A 1 {\displaystyle A_{1}\times A_{1}} A 2 {\displaystyle A_{2}} B 2 {\displaystyle B_{2}} C 2 {\displaystyle C_{2}} D 2 {\displaystyle D_{2}} G 2 {\displaystyle G_{2}}
3차원 기약 근계는 세 가지가 있으며, 이들은 정육면체 ·정팔면체 의 모양을 가진다.
반단순 리 대수에 대응되는 근계 복소수체 위의 반단순 리 대수 g {\displaystyle {\mathfrak {g}}} 및 그 카르탕 부분 대수 h ⊆ g {\displaystyle {\mathfrak {h}}\subseteq {\mathfrak {g}}} 가 주어졌다고 하자. h {\displaystyle {\mathfrak {h}}} 는 g {\displaystyle {\mathfrak {g}}} 의 킬링 형식 을 통해 자연스럽게 유한 차원 실수 내적 공간 을 이룬다.
그렇다면, g {\displaystyle {\mathfrak {g}}} 의 딸림표현 에 대응하는 h {\displaystyle {\mathfrak {h}}} -무게 들
Φ = { α ∈ h ∨ : g α ≠ 0 } ⊆ h ∨ {\displaystyle \Phi =\{\alpha \in {\mathfrak {h}}^{\vee }\colon {\mathfrak {g}}_{\alpha }\neq 0\}\subseteq {\mathfrak {h}}^{\vee }} 을 생각하자. 그렇다면, ( h ∨ , Φ ) {\displaystyle ({\mathfrak {h}}^{\vee },\Phi )} 는 근계를 이룬다. 또한, 다음이 성립한다.
g {\displaystyle {\mathfrak {g}}} 의 단순 리 대수 들로의 직합 분해는 ( h ∨ , Φ ) {\displaystyle ({\mathfrak {h}}^{\vee },\Phi )} 의 기약 근계들로의 직합 분해와 대응한다.특히, 단순 리 대수 에 대응하는 근계는 기약 근계이다. 두 반단순 리 대수 가 서로 동형일 필요 충분 조건 은 그 대응하는 근계가 서로 동형인 것이다.
역사
각주 Dynkin, E. B. The structure of semi-simple algebras. Uspehi Matem. Nauk (N.S.) 2, (1947). no. 4(20), 59–127.
외부 링크
같이 보기