diff --git a/docs/specifications/hdf5.Rmd b/docs/specifications/hdf5.Rmd
index 071485d..10df744 100644
--- a/docs/specifications/hdf5.Rmd
+++ b/docs/specifications/hdf5.Rmd
@@ -28,7 +28,7 @@ If not present, the version is assumed to be \"1.0\" for back-compatibility purp
     cat("This attribute should hold a scalar string dataset containing the value \"", as.character(.version), "\".", sep="")
 }
 ```
-This should use a HDF5 string datatype that is compatible with the UTF-8 encoding.
+This should use a datatype that can be represented by a UTF-8 encoded string.
 
 ```{r, echo=FALSE, results="asis"}
 if (.version >= package_version("1.3")) {
@@ -57,7 +57,7 @@ By providing unique names, users can improve interoperability with native data s
 An R list is represented as a HDF5 group (`**/`) with the following attributes:
 
 - `uzuki_object`, a scalar string dataset containing the value `"list"`.
-  This should use a HDF5 string datatype that is compatible with the UTF-8 encoding.
+  This should use a datatype that can be represented by a UTF-8 encoded string.
 
 This group should contain a subgroup `**/data` that contains the list elements.
 Each list element is itself represented by a subgroup that is named after its 0-based position in the list, e.g., `**/data/0` for the first list element.
@@ -65,14 +65,14 @@ One subgroup should be present for each integer in `[0, N)`, given a list of len
 Each list element may be any of the objects described in this specification, including further nested lists.
 
 If the list is named, there will additionally be a 1-dimensional `**/names` string dataset of length equal to the number of elements in `**/data`.
-This should use a HDF5 string datatype that is compatible with the UTF-8 encoding.
+This should use a datatype that can be represented by a UTF-8 encoded string.
 
 ### Atomic vectors
 
 An atomic vector is represented as a HDF5 group (`**/`) with the following attributes:
 
 - `uzuki_object`, a scalar string dataset containing the value `"vector"`.
-  This should use a HDF5 string datatype that is compatible with the UTF-8 encoding.
+  This should use a datatype that can be represented by a UTF-8 encoded string.
 ```{r, echo=FALSE, results="asis"}
 if (.version == package_version("1.0")) {
     cat('- `uzuki_type`, a scalar string dataset containing one of `"integer"`, `"boolean"`, `"number"`, `"string"`, `"date"` or `"date-time"`.')
@@ -80,7 +80,7 @@ if (.version == package_version("1.0")) {
     cat('- `uzuki_type`, a scalar string dataset containing one of `"integer"`, `"boolean"`, `"number"` or `"string"`.')
 }
 ```
-  This should use a HDF5 string datatype that is compatible with the UTF-8 encoding.
+  This should use a datatype that can be represented by a UTF-8 encoded string.
 
 The group should contain an 1-dimensional dataset at `**/data`.
 Vectors of length 1 may also be represented as a scalar dataset.
@@ -116,12 +116,12 @@ This should be a scalar string dataset that specifies constraints to the format
 - `"date"`: strings should be `YYYY-MM-DD` dates or the placeholder value.
 - `"date-time"`: strings should be in the Internet Date/Time format ([RFC 3339, Section 5.6](https://www.rfc-editor.org/rfc/rfc3339#section-5.6)) or the placeholder value.
 
-This should use a HDF5 string datatype that is compatible with the UTF-8 encoding.')
+This should use a datatype that can be represented by a UTF-8 encoded string.')
 }
 ```
 
 The atomic vector's group may also contain `**/names`, a 1-dimensional string dataset of length equal to that of `**/data`.
-This should use a HDF5 string datatype that is compatible with the UTF-8 encoding.
+This should use a datatype that can be represented by a UTF-8 encoded string.
 If `**/data` is a scalar, `**/names` should have length 1.
 
 ### Representing missing values
@@ -136,7 +136,7 @@ If no such attribute is present, it can be assumed that there are no missing val
 
 if (.version >= package_version("1.2")) {
     cat('The datatype of the placeholder attribute should be exactly the same as that of `**/data`, so as to avoid unexpected results upon casting.
-The only exception is when `**/data` is a string, in which case the placeholder may be of any string datatype that is compatible with the UTF-8 encoding.
+The only exception is when `**/data` is a string, in which case the placeholder may be of any string datatype that can be represented by a UTF-8 encoded string.
 it is expected that any comparison between the placeholder and strings in `**/data` will be performed bytewise in the same manner as `strcmp`.')
 }
 
@@ -170,7 +170,7 @@ If no such attribute is present, it can be assumed that there are no missing val
 A factor is represented as a HDF5 group (`**/`) with the following attributes:
 
 - `uzuki_object`, a scalar string dataset containing the value `"vector"`.
-  This should use a HDF5 string datatype that is compatible with the UTF-8 encoding.
+  This should use a datatype that can be represented by a UTF-8 encoded string.
 ```{r, echo=FALSE, results="asis"}
 if (.version == package_version("1.0")) {
     cat('- `uzuki_type`, a scalar string dataset containing `"factor"` or `"ordered"`.')
@@ -178,7 +178,7 @@ if (.version == package_version("1.0")) {
     cat('- `uzuki_type`, a scalar string dataset containing `"factor"`.')
 }
 ```
-  This should use a HDF5 string datatype that is compatible with the UTF-8 encoding.
+  This should use a datatype that can be represented by a UTF-8 encoded string.
 
 The group should contain an 1-dimensional dataset at `**/data`, containing 0-based indices into the levels.
 This should use a HDF5 integer datatype that can be represented by a 32-bit signed integer.
@@ -186,7 +186,7 @@ This should use a HDF5 integer datatype that can be represented by a 32-bit sign
 Missing values are represented as described above for atomic vectors.
 
 The group should contain `**/levels`, a 1-dimensional string dataset that contains the levels for the indices in `**/data`.
-This should use a HDF5 string datatype that is compatible with the UTF-8 encoding.
+This should use a datatype that can be represented by a UTF-8 encoded string.
 Values in `**/levels` should be unique.
 
 Values in `**/data` should be non-negative (missing values excepted) and less than the length of `**/levels`.
@@ -194,7 +194,7 @@ Note that the datatype constraints on `**/data` suggest that there should not be
 as beyond that, the levels cannot be indexed by elements of `**/data`.
 
 The group may also contain `**/names`, a 1-dimensional string dataset of length equal to `data`.
-This should use a HDF5 string datatype is compatible with the UTF-8 encoding.
+This should use a datatype that can be represented by a UTF-8 encoded string.
 
 ```{r, echo=FALSE, results="asis"}
 if (.version == package_version("1.1")) {
@@ -208,14 +208,14 @@ This should be interpreted as a boolean where a non-zero value specifies that we
 A "nothing" (a.k.a., "null", "none") value is represented as a HDF5 group with the following attributes:
 
 - `uzuki_object`, a scalar string dataset containing the value `"nothing"`.
-  This should use a HDF5 string datatype that is compatible with the UTF-8 encoding.
+  This should use a datatype that can be represented by a UTF-8 encoded string.
 
 ### External object
 
 Each external object is represented as a HDF5 group (`**/`) with the following attributes:
 
 - `uzuki_object`, a scalar string dataset containing the value `"external"`.
-  This should use a HDF5 string datatype that is compatible with the UTF-8 encoding.
+  This should use a datatype that can be represented by a UTF-8 encoded string.
 
 This should contain an `**/index` scalar dataset, containing an index that identifies this external object uniquely within the entire list.
 `**/index` should start at zero and be incremented whenever an external object is encountered. 
diff --git a/include/uzuki2/parse_hdf5.hpp b/include/uzuki2/parse_hdf5.hpp
index 6faf0d4..76aabde 100644
--- a/include/uzuki2/parse_hdf5.hpp
+++ b/include/uzuki2/parse_hdf5.hpp
@@ -91,7 +91,7 @@ void parse_integer_like(const H5::DataSet& handle, Host* ptr, Function check, co
 template<class Host, class Function>
 void parse_string_like(const H5::DataSet& handle, Host* ptr, Function check, hsize_t buffer_size) try {
     if (!ritsuko::hdf5::is_utf8_string(handle)) {
-        throw std::runtime_error("expected a string datatype with a UTF-8 compatible encoding");
+        throw std::runtime_error("expected a datatype that can be represented by a UTF-8 encoded string");
     }
 
     auto missingness = ritsuko::hdf5::open_and_load_optional_string_missing_placeholder(handle, "missing-value-placeholder");
@@ -178,7 +178,7 @@ void extract_names(const H5::Group& handle, Host* ptr, hsize_t buffer_size) try
 
     auto nhandle = handle.openDataSet("names");
     if (!ritsuko::hdf5::is_utf8_string(nhandle)) {
-        throw std::runtime_error("expected a string datatype with a UTF-8 compatible encoding");
+        throw std::runtime_error("expected a datatype that can be represented by a UTF-8 encoded string");
     }
 
     size_t len = ptr->size();
@@ -252,7 +252,7 @@ std::shared_ptr<Base> parse_inner(const H5::Group& handle, Externals& ext, const
         } else if (vector_type == "factor" || (version.equals(1, 0) && vector_type == "ordered")) {
             auto levhandle = ritsuko::hdf5::open_dataset(handle, "levels");
             if (!ritsuko::hdf5::is_utf8_string(levhandle)) {
-                throw std::runtime_error("expected a string datatype with a UTF-8 compatible encoding");
+                throw std::runtime_error("expected a datatype that can be represented by a UTF-8 encoded string");
             }
 
             int32_t levlen = ritsuko::hdf5::get_1d_length(levhandle.getSpace(), false);
@@ -300,7 +300,7 @@ std::shared_ptr<Base> parse_inner(const H5::Group& handle, Externals& ext, const
             } else if (handle.exists("format")) {
                 auto fhandle = check_scalar_dataset(handle, "format");
                 if (!ritsuko::hdf5::is_utf8_string(fhandle)) {
-                    throw std::runtime_error("expected a string datatype with a UTF-8 compatible encoding");
+                    throw std::runtime_error("expected a datatype that can be represented by a UTF-8 encoded string");
                 }
                 auto x = ritsuko::hdf5::load_scalar_string_dataset(fhandle);
                 if (x == "date") {
diff --git a/tests/src/date.cpp b/tests/src/date.cpp
index d0ba449..abcfd15 100644
--- a/tests/src/date.cpp
+++ b/tests/src/date.cpp
@@ -175,7 +175,7 @@ TEST(Hdf5DateTest, CheckError) {
         vhandle.createDataSet("format", H5::PredType::NATIVE_INT, H5S_SCALAR);
         create_dataset(vhandle, "data", { "harry", "ron", "hermoine" });
     }
-    expect_hdf5_error(path, "foo", "string datatype");
+    expect_hdf5_error(path, "foo", "can be represented by a UTF-8 encoded string");
 
     {
         H5::H5File handle(path, H5F_ACC_TRUNC);
diff --git a/tests/src/string.cpp b/tests/src/string.cpp
index 52d7588..fbeb8e0 100644
--- a/tests/src/string.cpp
+++ b/tests/src/string.cpp
@@ -164,7 +164,7 @@ TEST(Hdf5StringTest, CheckError) {
         auto ghandle = vector_opener(handle, "foo", "string");
         create_dataset<int>(ghandle, "data", { 1, 2, 3, 4, 5 }, H5::PredType::NATIVE_INT);
     }
-    expect_hdf5_error(path, "foo", "expected a string");
+    expect_hdf5_error(path, "foo", "UTF-8 encoded string");
 
     {
         H5::H5File handle(path, H5F_ACC_TRUNC);