The paper introduces Compositional Vector Grammars (CVGs), a parsing model that combines the speed of small-state Probabilistic Context-Free Grammars (PCFGs) with the semantic richness of neural word representations and compositional phrase vectors. CVGs use a recursive neural network (RNN) to learn syntactico-semantic, compositional vector representations for phrases and words, improving the Stanford Parser by 3.8% in terms of F1 score to 90.4%. The CVG model is trained in two stages: first, the base PCFG is trained, and then the syntactically untied RNNs are conditioned on the PCFG to learn the composition functions. This approach allows for different composition functions for different types of phrases, leading to significant improvements in parsing accuracy. The CVG model is also faster than the current Stanford factored parser, achieving a 20% speed improvement. The paper includes experiments on the Penn Treebank WSJ dataset, showing that the CVG model outperforms other parsers in terms of accuracy and speed.The paper introduces Compositional Vector Grammars (CVGs), a parsing model that combines the speed of small-state Probabilistic Context-Free Grammars (PCFGs) with the semantic richness of neural word representations and compositional phrase vectors. CVGs use a recursive neural network (RNN) to learn syntactico-semantic, compositional vector representations for phrases and words, improving the Stanford Parser by 3.8% in terms of F1 score to 90.4%. The CVG model is trained in two stages: first, the base PCFG is trained, and then the syntactically untied RNNs are conditioned on the PCFG to learn the composition functions. This approach allows for different composition functions for different types of phrases, leading to significant improvements in parsing accuracy. The CVG model is also faster than the current Stanford factored parser, achieving a 20% speed improvement. The paper includes experiments on the Penn Treebank WSJ dataset, showing that the CVG model outperforms other parsers in terms of accuracy and speed.