diff --git a/src/__tests__/parsers/pt9/interlinearXmlParser.test.ts b/src/__tests__/parsers/pt9/interlinearXmlParser.test.ts
index f2126ec7..1f4c6738 100644
--- a/src/__tests__/parsers/pt9/interlinearXmlParser.test.ts
+++ b/src/__tests__/parsers/pt9/interlinearXmlParser.test.ts
@@ -31,13 +31,11 @@ describe('InterlinearXmlParser', () => {
       `;
       const result = parser.parse(xml);
 
-      expect(result).toEqual({
-        ScrTextName: '',
+      expect(result).toStrictEqual({
         GlossLanguage: 'en',
         BookId: 'MAT',
         Verses: {
           'MAT 1:1': {
-            Hash: '',
             Clusters: [
               {
                 TextRange: { Index: 0, Length: 4 },
@@ -97,6 +95,40 @@ describe('InterlinearXmlParser', () => {
       expect(result.Verses['RUT 3:1'].Hash).toBe('123456');
     });
 
+    it('preserves the absent-vs-empty distinction for ScrTextName and Hash', () => {
+      const xml = `
+        <InterlinearData ScrTextName="" GlossLanguage="en" BookId="RUT">
+          <Verses>
+            <item>
+              <string>RUT 3:1</string>
+              <VerseData Hash="">
+                <Cluster>
+                  <Range Index="1" Length="2" />
+                  <Lexeme Id="x" />
+                </Cluster>
+              </VerseData>
+            </item>
+            <item>
+              <string>RUT 3:2</string>
+              <VerseData>
+                <Cluster>
+                  <Range Index="1" Length="2" />
+                  <Lexeme Id="x" />
+                </Cluster>
+              </VerseData>
+            </item>
+          </Verses>
+        </InterlinearData>
+      `;
+      const result = parser.parse(xml);
+
+      // Empty attribute values are preserved as empty strings, not dropped.
+      expect(result.ScrTextName).toBe('');
+      expect(result.Verses['RUT 3:1'].Hash).toBe('');
+      // An absent Hash attribute is PT9's not-approved state and must stay absent.
+      expect(result.Verses['RUT 3:2']).not.toHaveProperty('Hash');
+    });
+
     it('parses cluster with multiple lexemes and builds LexemesId and Id correctly', () => {
       const xml = `
         <InterlinearData GlossLanguage="en" BookId="MAT">
@@ -200,7 +232,7 @@ describe('InterlinearXmlParser', () => {
       expect(cluster.Id).toBe('10-3');
     });
 
-    it('parses Lexeme without GlossId as empty SenseId', () => {
+    it('parses Lexeme without GlossId as absent SenseId', () => {
       const xml = `
         <InterlinearData GlossLanguage="en" BookId="MAT">
           <Verses>
@@ -218,7 +250,30 @@ describe('InterlinearXmlParser', () => {
       `;
       const result = parser.parse(xml);
 
-      expect(result.Verses['MAT 1:1'].Clusters[0].Lexemes[0]).toEqual({
+      expect(result.Verses['MAT 1:1'].Clusters[0].Lexemes[0]).toStrictEqual({
+        LexemeId: 'Word:a',
+      });
+    });
+
+    it('preserves an empty GlossId attribute as an empty SenseId', () => {
+      const xml = `
+        <InterlinearData GlossLanguage="en" BookId="MAT">
+          <Verses>
+            <item>
+              <string>MAT 1:1</string>
+              <VerseData>
+                <Cluster>
+                  <Range Index="0" Length="1" />
+                  <Lexeme Id="Word:a" GlossId="" />
+                </Cluster>
+              </VerseData>
+            </item>
+          </Verses>
+        </InterlinearData>
+      `;
+      const result = parser.parse(xml);
+
+      expect(result.Verses['MAT 1:1'].Clusters[0].Lexemes[0]).toStrictEqual({
         LexemeId: 'Word:a',
         SenseId: '',
       });
@@ -284,7 +339,7 @@ describe('InterlinearXmlParser', () => {
       ]);
     });
 
-    it('omits Punctuation entries without valid Range', () => {
+    it('preserves Punctuation entries without a Range element, with no TextRange', () => {
       const xml = `
         <InterlinearData GlossLanguage="en" BookId="MAT">
           <Verses>
@@ -311,15 +366,20 @@ describe('InterlinearXmlParser', () => {
       `;
       const result = parser.parse(xml);
 
-      expect(result.Verses['MAT 1:1'].Punctuations).toHaveLength(1);
-      expect(result.Verses['MAT 1:1'].Punctuations[0]).toEqual({
-        TextRange: { Index: 1, Length: 2 },
-        BeforeText: 'c',
-        AfterText: 'd',
-      });
+      expect(result.Verses['MAT 1:1'].Punctuations).toStrictEqual([
+        {
+          BeforeText: 'a',
+          AfterText: 'b',
+        },
+        {
+          TextRange: { Index: 1, Length: 2 },
+          BeforeText: 'c',
+          AfterText: 'd',
+        },
+      ]);
     });
 
-    it('omits Punctuation entries when Range Index or Length is not finite (missing or non-numeric)', () => {
+    it('preserves Punctuation entries whose Range has missing or non-numeric attributes, with no TextRange', () => {
       const xml = `
         <InterlinearData GlossLanguage="en" BookId="MAT">
           <Verses>
@@ -357,12 +417,13 @@ describe('InterlinearXmlParser', () => {
       `;
       const result = parser.parse(xml);
 
-      expect(result.Verses['MAT 1:1'].Punctuations).toHaveLength(1);
-      expect(result.Verses['MAT 1:1'].Punctuations[0]).toEqual({
-        TextRange: { Index: 5, Length: 1 },
-        BeforeText: 'valid',
-        AfterText: '',
-      });
+      expect(result.Verses['MAT 1:1'].Punctuations).toStrictEqual([
+        { TextRange: { Index: 5, Length: 1 }, BeforeText: 'valid', AfterText: '' },
+        { BeforeText: 'no Index', AfterText: '' },
+        { BeforeText: 'no Length', AfterText: '' },
+        { BeforeText: 'non-numeric Index', AfterText: '' },
+        { BeforeText: 'non-numeric Length', AfterText: '' },
+      ]);
     });
 
     it('parses Punctuation with valid Range but missing BeforeText/AfterText as empty strings', () => {
@@ -426,7 +487,7 @@ describe('InterlinearXmlParser', () => {
       expect(result.Verses['MAT 1:2'].Clusters[0].Lexemes[0].LexemeId).toBe('b');
     });
 
-    it('parses item with missing VerseData as empty Hash, Clusters, Punctuations', () => {
+    it('parses item with missing VerseData as no Hash and empty Clusters, Punctuations', () => {
       const xml = `
         <InterlinearData GlossLanguage="en" BookId="MAT">
           <Verses>
@@ -438,8 +499,7 @@ describe('InterlinearXmlParser', () => {
       `;
       const result = parser.parse(xml);
 
-      expect(result.Verses['MAT 1:1']).toEqual({
-        Hash: '',
+      expect(result.Verses['MAT 1:1']).toStrictEqual({
         Clusters: [],
         Punctuations: [],
       });
@@ -458,8 +518,7 @@ describe('InterlinearXmlParser', () => {
       `;
       const result = parser.parse(xml);
 
-      expect(result.Verses['MAT 1:11']).toEqual({
-        Hash: '',
+      expect(result.Verses['MAT 1:11']).toStrictEqual({
         Clusters: [],
         Punctuations: [],
       });
@@ -594,7 +653,7 @@ describe('InterlinearXmlParser', () => {
 
       expect(result.GlossLanguage).toBe('en');
       expect(result.BookId).toBe('MAT');
-      expect(result.ScrTextName).toBe('');
+      expect(result.ScrTextName).toBeUndefined();
       expect(Object.keys(result.Verses).length).toBeGreaterThan(0);
 
       const mat11 = result.Verses['MAT 1:1'];
diff --git a/src/parsers/pt9/interlinearXmlParser.ts b/src/parsers/pt9/interlinearXmlParser.ts
index 8dbca4d4..ab9a8b2b 100644
--- a/src/parsers/pt9/interlinearXmlParser.ts
+++ b/src/parsers/pt9/interlinearXmlParser.ts
@@ -12,8 +12,12 @@ export interface StringRange {
 export interface LexemeData {
   /** ID of the lexeme (e.g. from Lexicon; XML attribute Id). */
   LexemeId: string;
-  /** ID of the sense/gloss used for this lexeme (XML attribute GlossId). */
-  SenseId: string;
+  /**
+   * ID of the selected sense (XML attribute GlossId, a sense id despite the historical name).
+   * Absent when the lexeme carries no sense selection; an empty attribute value is preserved as an
+   * empty string.
+   */
+  SenseId?: string;
 }
 
 /** Data on the interlinearization of a cluster. */
@@ -32,8 +36,11 @@ export interface ClusterData {
 
 /** Data on punctuation change. */
 export interface PunctuationData {
-  /** Character range this punctuation occupies in the verse text. */
-  TextRange: StringRange;
+  /**
+   * Character range this punctuation occupies in the verse text. Absent when the entry has no
+   * `Range` element with valid numeric attributes; the entry itself is still preserved.
+   */
+  TextRange?: StringRange;
   /** Punctuation text before the change (or empty). */
   BeforeText: string;
   /** Punctuation text after the change (or empty). */
@@ -42,8 +49,11 @@ export interface PunctuationData {
 
 /** Interlinear data for a single verse. */
 export interface VerseData {
-  /** Hash of verse text when approved; empty string if not approved. */
-  Hash: string;
+  /**
+   * Approval hash of the verse text. PT9 writes the attribute only when the verse is approved, so
+   * absence is the not-approved state; it is never coalesced to an empty string.
+   */
+  Hash?: string;
   /** Lexeme clusters in this verse. */
   Clusters: ClusterData[];
   /** Punctuation changes in this verse. */
@@ -52,8 +62,11 @@ export interface VerseData {
 
 /** Root interlinear data: book + verses. */
 export interface InterlinearData {
-  /** Source text / project name (e.g. from InterlinearData ScrTextName attribute). */
-  ScrTextName: string;
+  /**
+   * Source text / project name (legacy ScrTextName attribute, no longer written by modern PT9).
+   * Absent when the file does not carry the attribute.
+   */
+  ScrTextName?: string;
   /** Language code or name for the glosses. */
   GlossLanguage: string;
   /** Book id (e.g. "RUT", "MAT"). */
@@ -138,8 +151,8 @@ interface ParsedInterlinearXml {
 }
 
 /**
- * Maps a parsed Cluster's Lexeme children to {@link LexemeData}. A lexeme with no gloss id yields an
- * empty sense id rather than being dropped.
+ * Maps a parsed Cluster's Lexeme children to {@link LexemeData}. A lexeme with no GlossId yields an
+ * absent SenseId; an empty attribute value is preserved as an empty string.
  *
  * @throws {SyntaxError} If any Lexeme element is missing the required Id attribute.
  */
@@ -151,7 +164,8 @@ function extractLexemesFromCluster(clusterElement: ParsedCluster): LexemeData[]
     if (!lexemeId) {
       throw new SyntaxError('Invalid XML: Lexeme missing required Id attribute');
     }
-    return { LexemeId: lexemeId, SenseId: el['@_GlossId'] ?? '' };
+    const senseId = el['@_GlossId'];
+    return { LexemeId: lexemeId, ...(senseId !== undefined && { SenseId: senseId }) };
   });
 }
 
@@ -169,27 +183,24 @@ const parseStrictNumber = (raw: string | undefined): number | undefined => {
 /**
  * Maps a parsed VerseData's Punctuation array to {@link PunctuationData} array.
  *
- * Uses a **lenient** parsing strategy: entries without a valid `Range` (or with missing /
- * non-integer `Index`/`Length`) are silently filtered out rather than throwing. Punctuation data is
- * non-critical to the interlinear display; clusters are validated strictly in
+ * Every entry is preserved. `TextRange` is set only when the entry has a `Range` element with valid
+ * (non-negative integer) `Index`/`Length`; otherwise the entry keeps its texts with no range. PT9
+ * itself reads a missing `Range` as a `(0, 0)` default; absence is preserved here instead so no
+ * fabricated range enters the data. Clusters are validated strictly in
  * {@link extractClustersFromVerse} because they are required for alignment rendering.
  */
 function extractPunctuationsFromVerse(verseDataElement: ParsedVerseData): PunctuationData[] {
   const elements = verseDataElement.Punctuation ?? [];
 
-  return elements.flatMap((el) => {
-    const rangeElement = el.Range;
-    if (!rangeElement) return [];
-    const index = parseStrictNumber(rangeElement['@_Index']);
-    const length = parseStrictNumber(rangeElement['@_Length']);
-    if (index === undefined || length === undefined) return [];
-    return [
-      {
-        TextRange: { Index: index, Length: length },
-        BeforeText: el.BeforeText ?? '',
-        AfterText: el.AfterText ?? '',
-      },
-    ];
+  return elements.map((el) => {
+    const index = parseStrictNumber(el.Range?.['@_Index']);
+    const length = parseStrictNumber(el.Range?.['@_Length']);
+    return {
+      ...(index !== undefined &&
+        length !== undefined && { TextRange: { Index: index, Length: length } }),
+      BeforeText: el.BeforeText ?? '',
+      AfterText: el.AfterText ?? '',
+    };
   });
 }
 
@@ -245,8 +256,11 @@ function extractClustersFromVerse(verseDataElement: ParsedVerseData): ClusterDat
  * Parses interlinear XML strings into {@link InterlinearData} using fast-xml-parser.
  *
  * Input is a raw XML string (caller is responsible for obtaining it, e.g. from file or network).
- * Output is the {@link InterlinearData} shape declared in this module; no extra conversion is done.
- * Expects the interlinear XML schema described in [pt9-xml.md](pt9-xml.md).
+ * Output is the {@link InterlinearData} shape declared in this module, lossless with respect to
+ * optional data: absent attributes (`ScrTextName`, `Hash`, `GlossId`) stay absent rather than being
+ * coalesced to empty strings (`Hash` absence is PT9's not-approved state), and Punctuation entries
+ * without a valid `Range` are kept rangeless rather than dropped. Expects the interlinear XML
+ * schema described in [pt9-xml.md](pt9-xml.md).
  *
  * Each instance holds a configured `XMLParser`; create one parser and reuse it across multiple
  * `parse()` calls rather than constructing a new instance per file.
@@ -295,7 +309,7 @@ export class InterlinearXmlParser {
       throw new SyntaxError('Invalid XML: Missing InterlinearData root element');
     }
 
-    const scrTextName = root['@_ScrTextName'] ?? '';
+    const scrTextName = root['@_ScrTextName'];
     const glossLanguage = root['@_GlossLanguage'] ?? '';
     const bookId = root['@_BookId'] ?? '';
     if (!glossLanguage || !bookId) {
@@ -323,12 +337,13 @@ export class InterlinearXmlParser {
 
       const verseDataElement = item.VerseData;
       if (!verseDataElement) {
-        acc[verseKey] = { Hash: '', Clusters: [], Punctuations: [] };
+        acc[verseKey] = { Clusters: [], Punctuations: [] };
         return acc;
       }
 
+      const hash = verseDataElement['@_Hash'];
       acc[verseKey] = {
-        Hash: verseDataElement['@_Hash'] ?? '',
+        ...(hash !== undefined && { Hash: hash }),
         Clusters: extractClustersFromVerse(verseDataElement),
         Punctuations: extractPunctuationsFromVerse(verseDataElement),
       };
@@ -336,7 +351,7 @@ export class InterlinearXmlParser {
     }, {});
 
     return {
-      ScrTextName: scrTextName,
+      ...(scrTextName !== undefined && { ScrTextName: scrTextName }),
       GlossLanguage: glossLanguage,
       BookId: bookId,
       Verses: verses,
diff --git a/src/parsers/pt9/pt9-xml.md b/src/parsers/pt9/pt9-xml.md
index b570279e..d7896ebf 100644
--- a/src/parsers/pt9/pt9-xml.md
+++ b/src/parsers/pt9/pt9-xml.md
@@ -15,11 +15,11 @@ The extension reads PT9 interlinear data from XML files (e.g. `Interlinear_<lang
   - **Children:** Zero or more `item` elements. Each `item` represents one verse.
     - **`item`**
       - **`string`** (element text): Verse reference key (e.g. `"MAT 1:1"`, `"RUT 3:1"`). Must be unique in the document; duplicate references cause a parse error.
-      - **`VerseData`** (optional): If absent, the verse is stored with empty `Hash`, `Clusters`, and `Punctuations`.
+      - **`VerseData`** (optional): If absent, the verse is stored with no `Hash` and empty `Clusters` and `Punctuations`.
 
 - **VerseData**
   - **Attributes:**
-    - `Hash` (optional): Approval hash of the verse text when approved; empty if not approved.
+    - `Hash` (optional): Approval hash of the verse text. PT9 writes it only when the verse is approved, so absence is the not-approved state; the parser preserves absence (never coalesces to an empty string).
   - **Children:**
     - **`Cluster`** (zero or more): Word/morpheme clusters with range and lexemes.
     - **`Punctuation`** (zero or more): Punctuation change records.
@@ -31,23 +31,23 @@ The extension reads PT9 interlinear data from XML files (e.g. `Interlinear_<lang
     - **`Lexeme`** (zero or more): Lexemes in this cluster.
       - **Attributes:**
         - `Id` (required): Lexeme id (e.g. from a Lexicon).
-        - `GlossId` (optional): Sense/gloss id; omitted or empty is treated as empty string.
+        - `GlossId` (optional): Id of the selected sense (a sense id despite the historical attribute name). When absent, the parsed lexeme has no `SenseId`; an empty attribute value is preserved as an empty string.
     - **`Excluded`** (optional): Boolean flag indicating this instance of a phrase should be excluded from the interlinear display at this specific location. This is a very niche property that is included because it's possible to be present in the XML, even though it's rarely used. When `true`, the phrase is not displayed at this location but remains available elsewhere. The exclusion is location-specific (applies to this instance at this text range, not globally). Omitted or `false` means the phrase is included.
 
 - **Punctuation**
   - **Children:**
-    - **`Range`** (optional): If present, must have numeric `Index` and `Length`. Entries without a valid `Range` are skipped (not an error).
+    - **`Range`** (optional): Every Punctuation entry is preserved. `TextRange` is set only when `Range` is present with non-negative integer `Index` and `Length`; otherwise the entry has no `TextRange`. (PT9 itself reads a missing `Range` as a `(0, 0)` default; the parser preserves absence instead of fabricating a range.)
     - **`BeforeText`** (optional): Punctuation text before the change; omitted → empty string.
     - **`AfterText`** (optional): Punctuation text after the change; omitted → empty string.
 
 ## Parsed output (in-memory)
 
-The parser produces objects conforming to the types in `src/types/interlinearizer.d.ts`:
+The parser produces objects conforming to the types exported from `src/parsers/pt9/interlinearXmlParser.ts`. Optional data is preserved losslessly: absent XML attributes stay absent on the output objects rather than being coalesced to empty strings.
 
-- **InterlinearData:** `ScrTextName`, `GlossLanguage`, `BookId`, `Verses` (record of verse key → **VerseData**).
-- **VerseData:** `Hash`, `Clusters` (array of **ClusterData**), `Punctuations` (array of **PunctuationData**).
-- **ClusterData:** `TextRange` (`Index`, `Length`), `Lexemes` (array of `{ LexemeId, SenseId }`), `LexemesId` (slash-joined lexeme IDs), `Id` (cluster id: `LexemesId/Index-Length` or `Index-Length` when there are no lexemes), `Excluded` (boolean flag for location-specific exclusion).
-- **PunctuationData:** `TextRange`, `BeforeText`, `AfterText`.
+- **InterlinearData:** `ScrTextName?` (absent when the legacy attribute is missing), `GlossLanguage`, `BookId`, `Verses` (record of verse key → **VerseData**).
+- **VerseData:** `Hash?` (absent means the verse is not approved), `Clusters` (array of **ClusterData**), `Punctuations` (array of **PunctuationData**).
+- **ClusterData:** `TextRange` (`Index`, `Length`), `Lexemes` (array of `{ LexemeId, SenseId? }`), `LexemesId` (slash-joined lexeme IDs), `Id` (cluster id: `LexemesId/Index-Length` or `Index-Length` when there are no lexemes), `Excluded` (boolean flag for location-specific exclusion).
+- **PunctuationData:** `TextRange?` (absent when the entry has no valid `Range`), `BeforeText`, `AfterText`.
 
 ## Example (minimal valid document)