This section presents the key findings from the analysis of to-infinitive and -ing gerund constructions in PhE and SgE, based on the ICE-PHI and ICE-SIN corpora. The findings are structured into two main subsections, namely (1) distribution across spoken and written registers, and (2) verb-specific trends. Statistical analyses and tables provide empirical support for the patterns described.
3.3. Statistical Results: Verb Complementation Across Registers and Varieties
The statistical analysis using ANOVA validates the patterns observed in the distribution of to-infinitive and -ing gerund constructions across PhE and SgE.
Magpale and Ferrer (
2015) previously provided a frequency-based analysis of verb complementation in PhE; the current study extends this analysis by introducing normalized values and conducting a systematic ANOVA to assess the significance of register (spoken vs. written) and variety (PhE vs. SgE) in shaping verb complementation patterns. The analysis highlights key grammatical and sociolinguistic factors influencing the selection of to-infinitive and gerund complements, providing a clearer empirical basis for understanding the syntactic variability in Southeast Asian Englishes.
The effect size was measured using partial eta squared (η
2) to assess the proportion of variance explained by the main effects and interaction terms. Following
Cohen (
1988), η
2 values were interpreted as small (0.01), moderate (0.06), and large (0.14). This approach allows for a more precise evaluation of the strength of the observed patterns, enabling a nuanced understanding of the grammatical and sociolinguistic factors shaping verb complementation across registers and varieties.
3.3.1. Spoken Contexts
In spoken registers, to-infinitive constructions dominate in both PhE and SgE, but the ANOVA results (
Table 3a) reveal that the main effects of variety and verb type are not statistically significant, and there is no significant interaction between them. This suggests that the observed preference for to-infinitives in spoken discourse reflects a shared grammatical foundation rather than variety-specific differences.
The main effect of variety is not significant (F(1,28) = 0.071, p = 0.792) with a negligible effect size (η2 = 0.003), indicating that PhE and SgE exhibit similar overall patterns in spoken contexts. The main effect of verb type is statistically significant (F(1,28) = 7.56, p = 0.010) with a large effect size (η2 = 0.213), suggesting that while to-infinitives are preferred in both varieties, the difference is large enough to reach statistical significance. The interaction effect is non-significant (F(1,28) = 0.004, p = 0.952) with a negligible effect size (η2 = 0.000), implying that the influence of register on verb complementation is consistent across both PhE and SgE.
Despite the lack of statistical significance in the interaction term, the strong effect size for verb type indicates that the preference for to-infinitives in spoken contexts is robust. This aligns with the communicative demands of conversational discourse, where to-infinitives are favored for encoding intentionality and goal-directed actions. For example, in PhE, want shows a normalized frequency of 12.62 in spoken data, compared to 3.41 in written data, reflecting its role in expressing direct intention and goal setting in spontaneous, interactional contexts.
In SgE, the higher normalized frequency of to-infinitives in spoken contexts (e.g.,
want = 15.65) suggests a similar tendency toward direct goal expression. However, SgE demonstrates greater syntactic flexibility with gerunds. For instance,
like and
try occasionally select gerunds in SgE, suggesting that multilingual influence facilitates greater variation in spoken discourse. The large effect size for verb type (η
2 = 0.213) confirms that the preference for to-infinitives in spoken contexts is systematic and meaningful. The slightly greater syntactic flexibility in SgE, indicated by the higher use of gerunds, may reflect substrate influence from languages like Mandarin and Malay, where verb-complement structures are more fluid (
Bao, 2005).
3.3.2. Written Contexts
In written registers, the patterns reveal more structured preferences for verb complementation. The ANOVA results (
Table 3b) show a significant main effect of verb type (F(1,28) = 16.05,
p = 0.0004) with a large effect size (η
2 = 0.364), indicating that the difference between to-infinitives and gerunds is systematic and substantial in written contexts.
The non-significant effect for variety (F(1,28) = 0.082, p = 0.777) with a near-zero effect size (η2 = 0.003) suggests that both PhE and SgE demonstrate similar preferences for to-infinitives and gerunds in written contexts. The non-significant interaction effect (F(1,28) = 0.072, p = 0.790) with a negligible effect size (η2 = 0.003) further suggests that the distributional patterns of verb complementation are consistent across varieties in written discourse.
The significant main effect for verb type reflects a structural distinction between to-infinitives and gerunds in written registers. For instance, in PhE, enjoy shows a normalized frequency of 0.22 in written contexts, aligning with its semantic role in encoding habitual or descriptive actions. Similarly, in SgE, enjoy occurs with a normalized frequency of 0.14, reflecting similar grammatical constraints across varieties.
3.4. Syntactic Patterns of Subjectless Nonfinite Clauses in PhE and SgE
In
Section 3.1, the use of to-infinitive and -ing gerund constructions in PhE and SgE was explored, focusing on their frequency and distribution across spoken and written texts. Building on this,
Section 3.3 examines how subjectless nonfinite clauses function syntactically in PhE and SgE, specifically as direct objects in various types of clauses, including independent clauses, interrogative sentences, and relative clauses. PhE examples are sourced from
Magpale and Ferrer (
2015). This analysis, based on data from the International Corpus of English for the Philippines (ICE-PHI) and Singapore (ICE-SIN), provides further insights into how verb complementation in these varieties reflects sociocultural factors such as pragmatic strategies, discourse organization, and communicative context.
- a.
Direct Object in Independent Clauses
In both PhE and SgE, subjectless nonfinite clauses frequently function as direct objects of main verbs in independent clauses.
In (25), the infinitive clause
to sing away serves as the direct object of the verb
hopes. The construction specifies the intended action associated with the subject:
(25) | PhE: Dubbed as a senatoriable with a golden voice, she hopes to sing away to the hearts of the Filipinos with a campaign to fight graft and reduce taxes. (ICE-PHI:S1B-029#3:1:A) |
Similarly, in (26), the infinitive clause
to protect from data loss and virus attacks functions as the direct object of the verb
need to take, indicating a procedural requirement:
(26) | SgE: Organisations need to take preventive measures to protect from data loss and virus attacks, and also against computer frauds and breakdowns. (ICE-SIN:W2B-036#38:2) |
- b.
Direct Object in Interrogative Sentences
Subjectless nonfinite clauses also appear as direct objects in interrogative sentences, often embedded within verbs like want and hope.
In (27a) and (27b), the infinitive clause
to talk about functions as the direct object of the verb want, specifying the conversational focus:
(27a) | PhE: What do you want to talk about? (ICE-PHI:S1A-013#) |
(27b) | PhE: Why do you want to talk about Erap? (ICE-PHI:S1A-017#10:1:B) |
In SgE, similar patterns are observed, as in (28a) and (28b), where infinitive clauses act as direct objects to indicate future-oriented goals:
(28a) | SgE: Does it mean you want to go back to Malaysia and stay? (ICE-SIN:S1A-035#167:2:E) |
(28b) | SgE: How do you hope to go about achieving this? (ICE-SIN:S1B-049#51:1:D) |
- c.
Direct Object in Relative Clauses
In relative clauses, nonfinite clauses often serve as direct objects, providing additional specificity to the action or subject.
In (29), the infinitive clause
to eat functions as the direct object of want within the relative clause
who would want to eat sandwich, clarifying the subject’s preference:
(29) | PhE: It is my mother who would want to eat sandwich. (ICE-PHI:S1B-004#60:1:A) |
In (30), the infinitive clause
to write programs functions as the direct object of want, embedded in a technical context:
(30) | SgE: When the developers want to write programs that access the database, do they write the queries based on the conceptual model? (ICE-SIN:W2A-037#37:2) |
- d.
Direct Object in Hypothetical Statements
Nonfinite clauses often appear as direct objects in hypothetical constructions, indicating conditional or speculative scenarios.
In (31), the infinitive clause to join functions as the direct object of wants, expressing a conditional intention:
(31) | PhE: Well, if he wants to join us <&> speaker B chuckles. (ICE-PHI:S1A-037#138:1:A) |
Similarly, in (32), the infinitive clause
to build acts as the direct object of want, conveying a strategic proposal:
(32) | SgE: Okay, let’s say you want to build another country codes like uh Malaysia. (ICE-SIN:S2A-052#94:1) |
- e.
[Modal + Base Form] + Nonfinite Clause as Direct Object
Subjectless nonfinite clauses frequently occur as direct objects in modal constructions, emphasizing obligations or intentions.
In (33), the infinitive clause
to seek serves as the direct object of the verb phrase
would like, conveying a polite request:
(33) | PhE: Well uh right now we would like to seek the help of the government the D O T in seeking more liberal uh arrangements with uh countries. (ICE-PHI:S1B-027#82:1:G) |
In (34), the infinitive clause
to use leaded petrol functions as the direct object of need, aligning with procedural discourse:
(34) | SgE: The Ministry has assured the public that it will not be stopping the sale and use of leaded petrol since about ten percent of the cars in Singapore will still need to use leaded petrol. (ICE-SIN:S2B-017#20:1:B) |
- f.
[Negator + Main Verb] + Nonfinite Clause
Nonfinite clauses also serve as direct objects in negative constructions, where the negator modifies the main verb.
In (35), the infinitive clause
to delve acts as the direct object of wish, emphasizing the negation of an intended action:
(35) | PhE: I do not wish to delve into the issue of how and why almost two point five million Filipinos have migrated to North. (ICE-PHI:S2B-025#36:2:A) |
In (36), the infinitive clause to be functions as the direct object of
does not need, framing the absence of necessity:
(36) | SgE: A grammar test, therefore, does not need to be a discrete point test. (ICE-SIN:W2A-007#X108:1) |
Interestingly, the construction [Negator] + try + Nonfinite Clause functions as a syntactic mechanism for framing negated intentions or actions in both PhE and SgE. In these structures, the negator modifies the main verb try, while the nonfinite clause serves as its direct object, specifying the action being avoided.
- g.
[Negator] + try + Nonfinite Clause
This kind of construction serves as a syntactic mechanism for framing negated intentions or actions, with the negator modifying the main verb and the nonfinite clause functioning as its direct object.
In (37), the infinitive clause
to complain serves as the direct object of try not, highlighting a conscious effort to avoid an action:
(37) | PhE: But I I try <&> clears throat </&> I try not to complain too much at the restaurants you know whatever but because of ingesting other people ’s spit and so forth you know <&> speaker C laughs </&>. (ICE-PHI:S1B-035#27:1:B) |
The negator not modifies the main verb try, creating a syntactic split between the main verb and its complement. This construction emphasizes the speaker’s intentional restraint. The elaborative commentary following to complain aligns with PhE’s expressive tendency to provide detailed narratives and contextual elaboration. The frequent inclusion of hedging expressions (e.g., you know, whatever) further reflects a sociolinguistic preference for mitigating directness in negative constructions.
In (38), the infinitive clause to circumscribe acts as the direct object of
try not, similarly framing a deliberate avoidance of a specified action:
(38) | SgE: So I I try not to circumscribe lah otherwise it’ll be too much (ICE-SIN:S1A-034#231:1:C). |
Here, the negator not modifies try, and the infinitive clause to circumscribe specifies the negated behavior. The inclusion of the discourse particle lah reflects the pragmatic and interactive nature of SgE, softening the negation and situating it within a conversational framework.