diff --git a/base/docs/helpdb/Base.jl b/base/docs/helpdb/Base.jl
index db23f5bc621c7..f473b7ba402e2 100644
--- a/base/docs/helpdb/Base.jl
+++ b/base/docs/helpdb/Base.jl
@@ -9127,10 +9127,7 @@ on the `permute` and `scale` keyword arguments. The eigenvectors are returned co
 ```jldoctest
 julia> eig([1.0 0.0 0.0; 0.0 3.0 0.0; 0.0 0.0 18.0])
 ([1.0,3.0,18.0],
-3×3 Array{Float64,2}:
- 1.0  0.0  0.0
- 0.0  1.0  0.0
- 0.0  0.0  1.0)
+[1.0 0.0 0.0; 0.0 1.0 0.0; 0.0 0.0 1.0])
 ```
 
 `eig` is a wrapper around [`eigfact`](:func:`eigfact`), extracting all parts of the
diff --git a/doc/devdocs/reflection.rst b/doc/devdocs/reflection.rst
index a88c34d1236f7..185016495c732 100644
--- a/doc/devdocs/reflection.rst
+++ b/doc/devdocs/reflection.rst
@@ -92,7 +92,7 @@ the macro will be evaluated and the result will be passed instead!). For example
 .. doctest::
 
    julia> macroexpand( :(@edit println("")) )
-   :(Base.edit(println,(Base.typesof)("")))
+   :((Base.edit)(println,(Base.typesof)("")))
 
 The functions :func:`Base.Meta.show_sexpr` and :func:`dump` are used to display S-expr style views
 and depth-nested detail views for any expression.
diff --git a/doc/stdlib/linalg.rst b/doc/stdlib/linalg.rst
index bbf6ce5893a31..12d89f8e12875 100644
--- a/doc/stdlib/linalg.rst
+++ b/doc/stdlib/linalg.rst
@@ -584,10 +584,7 @@ Linear algebra functions in Julia are largely implemented by calling functions f
 
        julia> eig([1.0 0.0 0.0; 0.0 3.0 0.0; 0.0 0.0 18.0])
        ([1.0,3.0,18.0],
-       3×3 Array{Float64,2}:
-        1.0  0.0  0.0
-        0.0  1.0  0.0
-        0.0  0.0  1.0)
+       [1.0 0.0 0.0; 0.0 1.0 0.0; 0.0 0.0 1.0])
 
    ``eig`` is a wrapper around :func:`eigfact`\ , extracting all parts of the factorization to a tuple; where possible, using :func:`eigfact` is recommended.