diff --git a/src/bin/ali-to-pdf.cc b/src/bin/ali-to-pdf.cc
index 2b190853c94..61b5138cf31 100644
--- a/src/bin/ali-to-pdf.cc
+++ b/src/bin/ali-to-pdf.cc
@@ -34,7 +34,7 @@ int main(int argc, char *argv[]) {
         "Converts alignments (containing transition-ids) to pdf-ids, zero-based.\n"
         "Usage:  ali-to-pdf  [options] <model> <alignments-rspecifier> <pdfs-wspecifier>\n"
         "e.g.: \n"
-        " ali-to-pdf 1.mdl ark:1.ali ark, t:-\n";
+        " ali-to-pdf 1.mdl ark:1.ali ark,t:-\n";
     ParseOptions po(usage);
 
     po.Read(argc, argv);
diff --git a/src/doc/tree_externals.dox b/src/doc/tree_externals.dox
index df9f96e8430..01f8dc433bf 100644
--- a/src/doc/tree_externals.dox
+++ b/src/doc/tree_externals.dox
@@ -158,7 +158,7 @@ the questions configuration class would be output by the compile-questions
 program, which takes in a topology list of phonetic questions (in our
 scripts, these are automatically obtained from tree-building statistics
 by the program cluster-phones.  The roots file specifies sets of phones
-that are goint to have shared roots in the decision-tree clustering process, and says
+that are going to have shared roots in the decision-tree clustering process, and says
 for each phone set the following two things:
 
   - "shared" or "not-shared" says whether or not there should be separate roots