diff --git a/app/models/normalize_primo_record.rb b/app/models/normalize_primo_record.rb
index 35999f34..ba43cbfe 100644
--- a/app/models/normalize_primo_record.rb
+++ b/app/models/normalize_primo_record.rb
@@ -92,7 +92,7 @@ def year
   def format
     return unless @record['pnx']['display']['type']
 
-    normalize_type(@record['pnx']['display']['type'].join)
+    @record['pnx']['display']['type'].map { |term| Vocabularies::Format.lookup(term) }&.join(' ; ')
   end
 
   # While the links object in the Primo response often contains more than the Alma openurl, that is
@@ -277,15 +277,6 @@ def encode_author(author)
     URI.encode_uri_component(author)
   end
 
-  def normalize_type(type)
-    r_types = {
-      'BKSE' => 'eBook',
-      'reference_entry' => 'Reference Entry',
-      'Book_chapter' => 'Book Chapter'
-    }
-    r_types[type] || type.capitalize
-  end
-
   # It's possible we'll encounter records that use a different server,
   # so we want to test against our expected server to guard against
   # malformed URLs. This assumes all URL strings begin with https://.
diff --git a/app/models/normalize_timdex_record.rb b/app/models/normalize_timdex_record.rb
index 440fe951..9684a11c 100644
--- a/app/models/normalize_timdex_record.rb
+++ b/app/models/normalize_timdex_record.rb
@@ -102,10 +102,11 @@ def year
     end
   end
 
+  # This is the same as the content_type field below.
   def format
     return '' unless @record['contentType']
 
-    @record['contentType'].map { |type| type['value'] }.join(' ; ')
+    @record['contentType']&.map { |term| Vocabularies::Format.lookup(term) }&.join(' ; ')
   end
 
   def links
@@ -158,8 +159,10 @@ def identifier
   end
 
   # TIMDEX-specific methods
+
+  # This is the same as the format field above.
   def content_type
-    @record['contentType']
+    @record['contentType']&.map { |term| Vocabularies::Format.lookup(term) }&.join(' ; ')
   end
 
   def dates
diff --git a/app/models/vocabularies/format.rb b/app/models/vocabularies/format.rb
new file mode 100644
index 00000000..2018cfbb
--- /dev/null
+++ b/app/models/vocabularies/format.rb
@@ -0,0 +1,26 @@
+module Vocabularies
+  class Format
+    # FORMAT_MAPPINGS is an object listing all the machine-friendly format values we have encountered from TIMDEX or
+    # Primo, and the human-friendly values we want to normalize to. Entries should be alphabetized for easier
+    # maintenance.
+    FORMAT_MAPPINGS = {
+      'bkse' => 'eBook',
+      'book_chapter' => 'Book Chapter',
+      'conference_proceeding' => 'Conference Proceeding',
+      'magazinearticle' => 'Magazine Article',
+      'newsletterarticle' => 'Newsletter Article',
+      'reference_entry' => 'Reference Entry',
+      'researchdatabases' => 'Research Database'
+    }.freeze
+
+    # The lookup method attemps to look up a human-friendly value for any of the format values we get back from our
+    # source systems. The fetch method used allows a default value, which is what happens when a more human-friendly
+    # value isn't found in the FORMAT_MAPPINGS constant.
+    #
+    # @param value [String] A format value to be looked up, if a better version exists.
+    # @return [String, nil] The cleaned up version, or nil if a nil was submited.
+    def self.lookup(value)
+      FORMAT_MAPPINGS.fetch(value.downcase, value&.capitalize)
+    end
+  end
+end
diff --git a/app/views/search/_result.html.erb b/app/views/search/_result.html.erb
index d0f91bd3..ab39522e 100644
--- a/app/views/search/_result.html.erb
+++ b/app/views/search/_result.html.erb
@@ -11,7 +11,7 @@
     <div class="result-metadata">
 
       <p class="pub-info">
-        <span><%= result[:content_type]&.each { |type| type['value'] }&.join(' ; ') %></span>
+        <span><%= result[:content_type] %></span>
         
         <% if result[:date_range].present? %>
           <span><%= result[:date_range] %></span>
diff --git a/app/views/shared/_geo_data_info.html.erb b/app/views/shared/_geo_data_info.html.erb
index 74254731..5e7b9fe4 100644
--- a/app/views/shared/_geo_data_info.html.erb
+++ b/app/views/shared/_geo_data_info.html.erb
@@ -1,7 +1,8 @@
 <ul class="list-inline">
   <% if metadata[:content_type] %>
-    <li><%= metadata[:content_type]&.each { |type| type['value'] }&.join(' ; ') %></li>
+    <li><%= metadata[:content_type] %></li>
   <% elsif metadata['contentType'] %>
+    <% # This branch is used by the record view, which is not passed through a normalizer. %>
     <li><%= metadata['contentType']&.each { |type| type['value'] }&.join(' ; ') %></li>
   <% end %>
 
diff --git a/test/controllers/search_controller_test.rb b/test/controllers/search_controller_test.rb
index 330ef6f6..d0332447 100644
--- a/test/controllers/search_controller_test.rb
+++ b/test/controllers/search_controller_test.rb
@@ -95,7 +95,7 @@ def mock_timdex_search_success
       'api' => 'timdex',
       'title' => 'Sample TIMDEX Document Title',
       'timdexRecordId' => 'sample-record-123',
-      'contentType' => [{ 'value' => 'Article' }],
+      'contentType' => ['Article'],
       'dates' => [{ 'kind' => 'Publication date', 'value' => '2023' }],
       'contributors' => [{ 'value' => 'Foo Barston', 'kind' => 'Creator' }],
       'highlight' => [
@@ -143,7 +143,7 @@ def mock_timdex_search_all_tab
       'api' => 'timdex',
       'title' => 'Sample TIMDEX Document Title',
       'timdexRecordId' => 'sample-record-123',
-      'contentType' => [{ 'value' => 'Article' }],
+      'contentType' => ['Article'],
       'dates' => [{ 'kind' => 'Publication date', 'value' => '2023' }],
       'contributors' => [{ 'value' => 'Foo Barston', 'kind' => 'Creator' }],
       'highlight' => [
@@ -186,7 +186,7 @@ def mock_timdex_search_with_hits(total_hits)
       {
         'title' => "Sample TIMDEX Document Title #{i}",
         'timdexRecordId' => "sample-record-#{i}",
-        'contentType' => [{ 'value' => 'Article' }],
+        'contentType' => ['Article'],
         'dates' => [{ 'kind' => 'Publication date', 'value' => '2023' }],
         'contributors' => [{ 'value' => "Creator #{i}", 'kind' => 'Creator' }],
         'sourceLink' => "https://example.com/record#{i}"
@@ -448,7 +448,7 @@ def mock_timdex_search_with_hits(total_hits)
       'api' => 'timdex',
       'title' => 'Sample TIMDEX Document Title',
       'timdexRecordId' => 'sample-record-123',
-      'contentType' => [{ 'value' => 'Article' }],
+      'contentType' => ['Article'],
       'dates' => [{ 'kind' => 'Publication date', 'value' => '2023' }],
       'contributors' => [{ 'value' => 'Foo Barston', 'kind' => 'Creator' }],
       'highlight' => [],
diff --git a/test/fixtures/timdex/full_record.json b/test/fixtures/timdex/full_record.json
index 58927aa8..f9ee2f46 100644
--- a/test/fixtures/timdex/full_record.json
+++ b/test/fixtures/timdex/full_record.json
@@ -2,10 +2,7 @@
   "timdexRecordId": "test-record-123",
   "title": "Sample TIMDEX Record for Testing",
   "source": "Test Repository",
-  "contentType": [
-    {"value": "Dataset"},
-    {"value": "Geospatial data"}
-  ],
+  "contentType": ["Dataset", "Geospatial data"],
   "dates": [
     {"kind": "Publication date", "value": "2023-01-15"},
     {"kind": "Coverage", "value": "2020-2023"}
diff --git a/test/models/normalize_timdex_record_test.rb b/test/models/normalize_timdex_record_test.rb
index a166e797..4796364c 100644
--- a/test/models/normalize_timdex_record_test.rb
+++ b/test/models/normalize_timdex_record_test.rb
@@ -185,10 +185,7 @@ def minimal_record
   # Test TIMDEX-specific fields
   test 'includes TIMDEX-specific content_type field' do
     normalized = NormalizeTimdexRecord.new(full_record, 'test').normalize
-    expected_content_type = [
-      { 'value' => 'Dataset' },
-      { 'value' => 'Geospatial data' }
-    ]
+    expected_content_type = 'Dataset ; Geospatial data'
     assert_equal expected_content_type, normalized[:content_type]
   end
 
diff --git a/test/models/vocabularies/format_test.rb b/test/models/vocabularies/format_test.rb
new file mode 100644
index 00000000..ceff7daf
--- /dev/null
+++ b/test/models/vocabularies/format_test.rb
@@ -0,0 +1,15 @@
+require 'test_helper'
+
+class VocabularyFormatTest < ActiveSupport::TestCase
+  test 'lookup method returns better values where we know them' do
+    value = 'BKSE'
+    output = Vocabularies::Format.lookup(value)
+    assert_equal output, 'eBook'
+  end
+
+  test 'lookup method returns sentence case as a default' do
+    value = 'UNEXPECTED VALUE'
+    output = Vocabularies::Format.lookup(value)
+    assert_equal output, 'Unexpected value'
+  end
+end